AR3 woofer VC DCR wacko?

[Mazeppa]

Greetings fellow ARistas. I've decided to salvage the usable drivers out of a pair of AR3 that both have bad mids. I'm going to concentrate my efforts on a pair of AR5.

So, I hook the AR3s up for a last check prior to disassembly, yep woofs and tweets singing right along, both mids dead. Sound surprisingly good, actually. Pulled the woofers & tweets (Yes, I used a Dremel cut off wheel to cut the tabs by their rivets the aluminum wires are soldered to. Hey, I read classicspeakerpages so I know these things.) Started double checking VC DCR on the drivers; mids are open, tweets are both 1.7 ohm, woofers = 1.5-4.2-1.5-4.2 ohm repeat ad nauseum. Hooked woofers up to a source, playing just fine. Check VC DCR again, repeat previous readings. I expected a nice steady reading, but noooooo!

I was going to offer these drivers to forum members first in case somebody needed them, but I don't want to be pushing bogus goods. Anybody out there have the explain for this phenomenom?

BTW, woofers are the cast aluminum basket, rubberized cloth surround types which I understand to be the original types

[tysontom]

Greetings fellow ARistas. I've decided to salvage the usable drivers out of a pair of AR3 that both have bad mids. I'm going to concentrate my efforts on a pair of AR5.

So, I hook the AR3s up for a last check prior to disassembly, yep woofs and tweets singing right along, both mids dead. Sound surprisingly good, actually. Pulled the woofers & tweets (Yes, I used a Dremel cut off wheel to cut the tabs by their rivets the aluminum wires are soldered to. Hey, I read classicspeakerpages so I know these things.) Started double checking VC DCR on the drivers; mids are open, tweets are both 1.7 ohm, woofers = 1.5-4.2-1.5-4.2 ohm repeat ad nauseum. Hooked woofers up to a source, playing just fine. Check VC DCR again, repeat previous readings. I expected a nice steady reading, but noooooo!

I was going to offer these drivers to forum members first in case somebody needed them, but I don't want to be pushing bogus goods. Anybody out there have the explain for this phenomenom?

BTW, woofers are the cast aluminum basket, rubberized cloth surround types which I understand to be the original types

Mazeppa,

First of all, you did not need to cut the tabs on the tweeter terminal strip; all you had to do was to apply a soldering gun with high heat, and unravel the coils of aluminum wire to free the drivers. The dcr on your tweeters is about in spec; I don't understand what you mean about the woofers. Did you mean that one was 1.5 ohms and the other was 4.2 ohms, or was the value drifting? The first part seems plausible. The higher reading unit should be in the 2-3 ohm range and the lower reading is okay but a bit low. One or both woofers might have received a heavy dose of amplifier power over a long period of time, but they may be okay. Always be certain to free the drivers completely from any crossover components before measuring, but it sounds as though you did this when you measured them.

You might want to pry off the protective screen over the midrange driver and take a look at the two aluminum wires that loop up slightly (a sort of "Shephard's hook") and then go into the side of the dome close to the outside edge. More often than not, the aluminum wire would fatigue and actually break at that junction, causing an "open" circuit. With great patience, you can actually repair this break -- that is if you feel up to the task. Aluminum is hard to solder, and the union at that point is very difficult. If there are not breaks in the connection, then the voice coil was lost to a thermal overload.

--Tom Tyson

[soundminded]

Greetings fellow ARistas. I've decided to salvage the usable drivers out of a pair of AR3 that both have bad mids. I'm going to concentrate my efforts on a pair of AR5.

So, I hook the AR3s up for a last check prior to disassembly, yep woofs and tweets singing right along, both mids dead. Sound surprisingly good, actually. Pulled the woofers & tweets (Yes, I used a Dremel cut off wheel to cut the tabs by their rivets the aluminum wires are soldered to. Hey, I read classicspeakerpages so I know these things.) Started double checking VC DCR on the drivers; mids are open, tweets are both 1.7 ohm, woofers = 1.5-4.2-1.5-4.2 ohm repeat ad nauseum. Hooked woofers up to a source, playing just fine. Check VC DCR again, repeat previous readings. I expected a nice steady reading, but noooooo!

I was going to offer these drivers to forum members first in case somebody needed them, but I don't want to be pushing bogus goods. Anybody out there have the explain for this phenomenom?

BTW, woofers are the cast aluminum basket, rubberized cloth surround types which I understand to be the original types

Check the batteries in your meter. Try another meter. The only explanation I can think of that would result in this oscillating measurement involving the woofers is an intermittent internal short between VC windings caused by applying the meter voltage itself but that voltage is very low and if that were the case, the woofers wouldn't function properly. The chances that this happened to even one of them let alone identically to both of them is very slim. I'll put my money on the meter not functioning properly.

[Mazeppa]

Mazeppa,

First of all, you did not need to cut the tabs on the tweeter terminal strip; all you had to do was to apply a soldering gun with high heat, and unravel the coils of aluminum wire to free the drivers. The dcr on your tweeters is about in spec; I don't understand what you mean about the woofers. Did you mean that one was 1.5 ohms and the other was 4.2 ohms, or was the value drifting? The first part seems plausible. The higher reading unit should be in the 2-3 ohm range and the lower reading is okay but a bit low. One or both woofers might have received a heavy dose of amplifier power over a long period of time, but they may be okay. Always be certain to free the drivers completely from any crossover components before measuring, but it sounds as though you did this when you measured them.

Yeah, both woofers were out of circuit, sitting on the bench. Readings for both woofers were drifting up and down through approx. the range listed.

You might want to pry off the protective screen over the midrange driver and take a look at the two aluminum wires that loop up slightly (a sort of "Shephard's hook") and then go into the side of the dome close to the outside edge. More often than not, the aluminum wire would fatigue and actually break at that junction, causing an "open" circuit. With great patience, you can actually repair this break -- that is if you feel up to the task. Aluminum is hard to solder, and the union at that point is very difficult. If there are not breaks in the connection, then the voice coil was lost to a thermal overload.

--Tom Tyson

Never heard this one. Nothing to lose in trying. Thanks for the tip.

[Mazeppa]

Check the batteries in your meter. Try another meter.

No low batt. indications. Should have mentioned I did try 2 different meters, same result from both.

The only explanation I can think of that would result in this oscillating measurement involving the woofers is an intermittent internal short between VC windings caused by applying the meter voltage itself but that voltage is very low and if that were the case, the woofers wouldn't function properly. The chances that this happened to even one of them let alone identically to both of them is very slim.

My thoughts exactly. Maybe one, but both?

I'll put my money on the meter not functioning properly.

I'll try w/fresh batteries when I get home this evening.

[genek]

I was never able to get good DCR measurements from old speakers until I got myself a set of test leads with spring clips on their working ends. Old speaker terminals seem to develop some sort of oxide coating that makes it hard to get reliable electrical contact between them and anything that isn't soldered on or clamped down good and tight.

[soundminded]

I was never able to get good DCR measurements from old speakers until I got myself a set of test leads with spring clips on their working ends. Old speaker terminals seem to develop some sort of oxide coating that makes it hard to get reliable electrical contact between them and anything that isn't soldered on or clamped down good and tight.

Highly accurate readings of very low impedence loads or where high precision is required is (was) often done with a Wheatstone bridge using precision resistors. The correct value is obtained when the bridge is in perfect balance. Impedence of the windings will also change with temperature but nealy as much as was reported here especially with the low current a meter uses. A good secure connection is also required so any film that would add resistance will throw the readings off. Another method is to use a voltage source, in this case a battery, a volt meter, and and in line ammeter.

If the batteries in the meter are weak, the act of making the measurement can draw the battery voltage down causing the reading to change. If the lead is momentarily removed, the battery may recover to repeat the process for a considerable number of times until the battery is dead. Also the battery terminals in the meter and the contacts in the meter switch should be checked and cleaned. A battery terminal with oxide from an old corroded battery that wasn't cleaned properly can cause erratic readings. So can oxide on a meter function switch. Check the meter calibration by shorting the leads together. If you get more than 0.1 or 0.2 ohms, the battery is probably very weak.

[tysontom]

Greetings fellow ARistas. I've decided to salvage the usable drivers out of a pair of AR3 that both have bad mids. I'm going to concentrate my efforts on a pair of AR5.

So, I hook the AR3s up for a last check prior to disassembly, yep woofs and tweets singing right along, both mids dead. Sound surprisingly good, actually. Pulled the woofers & tweets (Yes, I used a Dremel cut off wheel to cut the tabs by their rivets the aluminum wires are soldered to. Hey, I read classicspeakerpages so I know these things.) Started double checking VC DCR on the drivers; mids are open, tweets are both 1.7 ohm, woofers = 1.5-4.2-1.5-4.2 ohm repeat ad nauseum. Hooked woofers up to a source, playing just fine. Check VC DCR again, repeat previous readings. I expected a nice steady reading, but noooooo!

I was going to offer these drivers to forum members first in case somebody needed them, but I don't want to be pushing bogus goods. Anybody out there have the explain for this phenomenom?

BTW, woofers are the cast aluminum basket, rubberized cloth surround types which I understand to be the original types

You may have mentioned this somewhere, but what type of meter were you using? You really must use a high-quality digital voltmeter or precision VTVM (e.g., calibrated HP) to get accurate results for this type of very low-resistance measurement. Using a Wheatstone bridge is over-kill, in my view, but it would be even more accurate. A good Fluke digital meter is probably best overall method for this sort of thing, and. This drifting-resistance phenomenon is uncommon and seems to be related to the meter rather than the voice coils.

--Tom tyson

[genek]

If the batteries in the meter are weak, the act of making the measurement can draw the battery voltage down causing the reading to change. If the lead is momentarily removed, the battery may recover to repeat the process for a considerable number of times until the battery is dead. Also the battery terminals in the meter and the contacts in the meter switch should be checked and cleaned. A battery terminal with oxide from an old corroded battery that wasn't cleaned properly can cause erratic readings. So can oxide on a meter function switch. Check the meter calibration by shorting the leads together. If you get more than 0.1 or 0.2 ohms, the battery is probably very weak.

In the olden days I had a Simpson VTVM, and I could tell that I was getting false "low" readings due to intermittently failing test lead contact because the meter needle would bounce around like a pebble in an earthquake. Today I have access to an electronics lab that is equipped with Wheatstone and double-Kelvin bridge testers, but for home use am flying with a cheapo DMM I got for four bucks at Harbor Freight, which I mostly just use to verify continuity. However, I took it to the lab and did some comparisons just for laughs, and as long as I have fresh batteries in the HF toy and use my spring clip leads, I can usually depend on the toy to be within +- .25 ohm in the 1-15 ohm range. If I really need to know what the exact impedance of something is, I take it to the lab. It's actually quite amazing comparing what you can get today for pocket change to what was expensive 30 years ago.

[Pete B]

Greetings fellow ARistas. I've decided to salvage the usable drivers out of a pair of AR3 that both have bad mids. I'm going to concentrate my efforts on a pair of AR5.

So, I hook the AR3s up for a last check prior to disassembly, yep woofs and tweets singing right along, both mids dead. Sound surprisingly good, actually. Pulled the woofers & tweets (Yes, I used a Dremel cut off wheel to cut the tabs by their rivets the aluminum wires are soldered to. Hey, I read classicspeakerpages so I know these things.) Started double checking VC DCR on the drivers; mids are open, tweets are both 1.7 ohm, woofers = 1.5-4.2-1.5-4.2 ohm repeat ad nauseum. Hooked woofers up to a source, playing just fine. Check VC DCR again, repeat previous readings. I expected a nice steady reading, but noooooo!

I was going to offer these drivers to forum members first in case somebody needed them, but I don't want to be pushing bogus goods. Anybody out there have the explain for this phenomenom?

BTW, woofers are the cast aluminum basket, rubberized cloth surround types which I understand to be the original types

Have you made sure that the cones are remaining still?

Any motion will produce current in the VC and completely

upset the measurement. I sometimes put some damping

between the cone and the basket to reduce any small motion.

[Mazeppa]

Fresh batteries in the Fluke 83 Series III resulted in steady readings of 2.6 ohm and 2.7 ohm from the 'ol AR3 woofers.

I must say I'm a little embarrassed to have neglected such an obvious thing to try. I'm glad to have the suggestion from Tom Tyson about the possibility of repair for the mids, will definitely give that a try.

Thanks for the help, guys.

[Mazeppa]

You might want to pry off the protective screen over the midrange driver and take a look at the two aluminum wires that loop up slightly (a sort of "Shephard's hook") and then go into the side of the dome close to the outside edge. More often than not, the aluminum wire would fatigue and actually break at that junction, causing an "open" circuit. With great patience, you can actually repair this break -- that is if you feel up to the task. Aluminum is hard to solder, and the union at that point is very difficult. If there are not breaks in the connection, then the voice coil was lost to a thermal overload.

--Tom Tyson

Pried the screens off of my afflicted AR3 mids. Found one pair of leads to be intact, still attached and entering into the dome area, so the open is somewhere inside. On the other one a lead was indeed open on the outside, right where it enters the dome area with barely enough protruding to grasp with a tweeser. Doesn't seem to be enough conductor accessible to affect a repair.

So, both of mine are a wash, but anyone having troubled mids, it's certainly worht a look in there, you just mught be able to get 'em back online.

I have a friend who is in possession of a pair of AR3a with both mids open. Could this procedure perhaps be applicable?

[tysontom]

Pried the screens off of my afflicted AR3 mids. Found one pair of leads to be intact, still attached and entering into the dome area, so the open is somewhere inside. On the other one a lead was indeed open on the outside, right where it enters the dome area with barely enough protruding to grasp with a tweeser. Doesn't seem to be enough conductor accessible to affect a repair.

So, both of mine are a wash, but anyone having troubled mids, it's certainly worht a look in there, you just mught be able to get 'em back online.

I have a friend who is in possession of a pair of AR3a with both mids open. Could this procedure perhaps be applicable?

Mazeppa,

It's good that you attempted the leadwire-repair routine. You sometimes have to cut very slightly into the phenolic-dome portion to get a little more of the wire exposed; then you can make a slight bend in it (a "hook" of sorts) to grip the in-coming wire, then try to crimp the ends together and try also to solder -- even with conventional solder -- the connection. This takes great patience, but I have done it two or three times with success. I doubt the "repaired" dome would have passed frequency-response QC tests at AR back in the 60s, but at least the dome was back in service. But since you have another dome that appears to have a thermal burn-out (actually fairly rare with this driver), you might just be better off getting another set of dome-midrange drivers. If you find some AR-3a midrange drivers, you can remove the protective screen and the fiberglass, and leave the dome exposed. There is a crossover modification you can make (see the AR-3a Refurbishing article in these pages) to the AR-3 to make it compatible with this driver. Alternatively, find a couple of eBay AR-3 2-inch midranges if you can.

This type of failure doesn't happen to the AR-3a dome -- the connections are very different, and the AR-3a also uses copper wire for both the midrange and the tweeter vs. the aluminum wire for the AR-3. Copper is less brittle than aluminum, and can sustain any stresses better.

--Tom Tyson

[speaker dave]

Just a thought on the problem of soldering aluminum wire. Don't forget that there are conductive paints available. I bought a small bottle from an electronics supply house. I needed it to bridge across a small break in a volume pot that I couldn't find a replacement for.

You should be able to strip a short bit of both ends of the wire (scrape bare with a razor blade), twist them together and then seal with a drop of the conductive paint. Worth a try.

David

[tysontom]

Just a thought on the problem of soldering aluminum wire. Don't forget that there are conductive paints available. I bought a small bottle from an electronics supply house. I needed it to bridge across a small break in a volume pot that I couldn't find a replacement for.

You should be able to strip a short bit of both ends of the wire (scrape bare with a razor blade), twist them together and then seal with a drop of the conductive paint. Worth a try.

David

David,

Yes, I have heard (was it John O'Hanlon?) about "gluing" the aluminum. The problem with the break-at-the-dome location is the lack of sufficient aluminum wire to attach; twisting together is usually out of the question, and the aluminum will usually break with twisting. In addition, the stress-break for an AR dome usually occurs a millimeter or two directly away from the phenolic material itself. Making a small "hook" on both ends, then crimping together and soldering or gluing will sometime work if enough material can be found to connect. Aluminum can definitely be soldered with aluminum solder, but it is tedious and difficult, the ever-present but "never-ending battle against the forces of evil."

Aluminum, formerless, voice coils and huge magnets in the AR-3 made for good performance, but it also made for difficult repair -- sometimes impossible. I don't believe the AR factory ever wasted time on "repairing" a dome tweeter from the AR-3 series; rather, the magnet assembly was stripped and a new dome applied or simply the structure was scrapped altogether in lieu of a new tweeter.

--Tom Tyson