• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Deviation in DC Resistance/ Ohm, tweeter. How much is acceptable?

Verdinut

Member
Joined
Mar 21, 2020
Messages
24
Likes
24
Location
Montreal, Canada
Actually, this is not correct. I even described how it happens (twice), perhaps you didn't read it. I saw other posters mention here too. Let me explain it again but maybe you won't read it anyway.
If a voice coil is overheated, two or more adjacent windings on the voice coil can short. These inter-winding shorts, the shorted windings no longer plays a role and the rest of the VC remains conductive. The reduction in resistance depends on the quantity of coils that got fused to each other. If 7% of the VC windings are shorted, the DCR reduces by ~7%, and the T/S parameters for the driver dramatically change (all of those shorted windings no longer play a role in the motor, inductance drops, and the defect introduces a non-linearity into the motor because the magnetic field in the vicinity of the short is quite non-uniform. Lots of stuff moves in this case, and the driver ends up testing much different than a good driver for the electrical parameters associated with the motor. Different drivers have different probability of this type of fail, some always blew to an open circuit, some have a large probability of the VC developing a partial short due to abuse. I have seen a few drivers with partial shorts due to misaligned VC rubbing the magnet.
And in case you don't want to read my text, here is a tutorial on voice coils that might help you visualize what is going on, and it touches on the problem of shorted turns, although more from the perspective of the type of bobbin the coil is wound on.

Yes I have. Perhaps you noticed I tested 22 drivers lying around my shop in this thread! It's not an accident! Back in the day, I installed thousands of car stereos, most built from separates. I worked at home hifi stores too, just slightly less blown tweeters! I can typically tell just from the sound if a driver has a VC rub or is partially shorted, so could all my co-workers. One thing for sure, if I got handed a pair of tweeter from the parts counter that were mismatched by 5%, no way I would begin the install... We tested every driver on an Ohm meter before install, no way you want to find out at the end of the build that you got a bad driver. So, yeah, experience. Literally thousands of drivers. YOu just don't see large DCR on a regular basis unless you measured wrong.

Perhaps you didn't abuse your tweeters the way some of my customers did! Interesting about your Altec Lansing drivers, I really don't know what to say about such a large mismatch without the data. Voice Coils are a length of wire wrapped into a precise number of turns. It is extremely unusual to see variation much larger than 1%, wire resistivity between samples, and hard to imagine the factory accidentally wound 25% more turns on the bobbin. It is very challenging to get the number of turns or the resistivity of the wire that different. In reality, VC matching is really quite precise, and not subject to large variation on a speaker that is working properly. That's why voice coil DCR variation is typically closer to 1% as I showed, not 7% as the OP measured, and certainly not 25% like you saw.
Well, it gets pretty hard to abuse at home pro Altec drivers when they have a sensitivity at 4 feet ranging from 99 dB to 108 dB for a 2.83 volt input. A pair of A7 VOTT speaker drivers in an appropriate cabinet can take a peak power of 700 watts! Just use an overall sensitivity of 100 dB (103 dB for a pair) and you would get peaks of over 125 dB at an 8 foot distance which are ear damaging levels.
 
Last edited:

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,135
Likes
4,779
Location
Portland, OR, USA
Well, it gets pretty hard to abuse at home pro Altec drivers when they have a sensitivity at 4 feet ranging from 99 dB to 108 dB for a 2.83 volt input. A pair of A7 VOTT speaker drivers in an appropriate cabinet can take a peak power of 700 watts! Just use an overall sensitivity of 100 dB (103 dB for a pair) and you would get peaks of over 125 dB at an 8 foot distance which are ear damaging levels.
Yeah. Those VOTT do tend to avoid getting blown by virtue of efficiency and power capability. Real classics.

Elliott Sound Products has a great page on speaker failure modes that I forgot about. It's a really useful site, I bought and built some of his electronics.
Section 7 has a description of the various voice coil failure modes, starting with the partially blown.
7 - Burnt Speaker Forensics
A burnt voice coil often tells a story, much like the evidence left at a crime scene does. The degree of burning or damage seen is not conclusive because the initial damage that led to failure can be very minor but turn into major damage when amplifier power continues to be delivered, as it normally does, until sound output stops.

A few examples will help illustrate the idea:

1. Black discolouration in the middle or all over the voice coil
This is the most common damage seen and indicates that a well centred coil was driven with too much audio power for too long. The resistance of the burnt coil is often half or less the nominal value due to internal shorting.

2. Black discolouration at one end of the voice coil.
This is also a common sight and indicates either the voice coil was not centred during manufacture OR that the damage was caused by DC current rather than audio frequency current. A large DC current will displace the voice coil to one or other extreme. A faulty amplifier is automatically suspected.

3. Black discolouration at both ends of the voice coil
More likely to be seen with large excursion and 'Hi-Fi' sub woofers where the voice coil is longer than the magnet gap depth. The portions that 'overhang' the gap are not so well cooled and will burn up first.

4. Scraping marks and black spots on the voice coil
This is a sure sign of metal particles caught in the gap. The most likely time for such particles to enter is when a speaker is being re-coned and the gap is wide open after the old cone and voicecoil have been removed.

5. Loose wires hanging off the voice coil
Typical of a adhesive failure at high temperature. The adhesive used may have been of low temp grade or was not correctly mixed.

6. Looks fine but tests open circuit
This is a nasty one as it indicates bad manufacture. The ends of the voice coil wire were not terminated well enough for the speaker to survive normal use. Most often seen where anodised Aluminium wire or strip is used for the voice coil. This failure is seen with horn diaphragms too.
spk-fail-f6.jpg
 

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,607
Likes
4,514
Location
Germany
Just think how the DC resistance can change,
1) Material of voicecoil wire
2) Diameter of voicecoil wire
3) Length of voicecoil wire.
4) temperatur
Thats it.

Lets imagine temperatur was the same.

Changing 1,2,3 will change magnetic field and this will lead to changed accustic output if you not change magnet or ...,,

10% of change in DC resistance is a lot and i dont think any modern driver manufacturer that wants be on top can,would accept this.

At least my ten cent
So measurement failure, Voice coil failure is more plausible. Or they changed the complete driver in many aspects to keep same accustic output and keep the same naming. What would also be not nice.

To make it more easy to understand its like using 100 or 110 windings on the coil. And a winding machine thats 10 windings wrong out of 100 has realy bad problems. Than you better let count a 6 year old. ;)

So from my point of view 10% of DC voice coil variance are a absolute no go for a quality driver manufacturer.
Let me make some fun.

Customer: you told me that the right and left speaker are the same?

High end Seller: yes sure

Customer: But one is 1m and the other 1,10m ?

High end seller: oh, thats tolarence has no influence in what you hear.

Costumer: But i hear a difference?

High end seller: you use old copper cables, thats the problem, get this new silver cables and they will sound the same.

;)
 
Last edited:
OP
DanielT

DanielT

Major Contributor
Joined
Oct 10, 2020
Messages
4,751
Likes
4,633
Location
Sweden - Слава Україні
Just think how the DC resistance can change,
1) Material of voicecoil wire
2) Diameter of voicecoil wire
3) Length of voicecoil wire.
4) temperatur
Thats it. If you like the same acoustic output,

Lets imagine temperatur was the same.

Changing 1,2,3 will change magnetic field and this will lead to changed accustic output if you not change magnet or ...,,

10% of change in DC resistance is a lot and i dont think any modern driver manufacturer that wants be on top can,would accept this.

At least my ten cent
So measurement failure, Voice coil failure is more plausible. Or they changed the complete driver in many aspects to keep same accustic output and keep the same naming. What would also be not nice.

To make it more easy to understand its like using 100 or 110 windings on the coil. And a winding machine thats 10 windings wrong out of 100 has realy bad problems. Than you better let count a 6 year old. ;)

So from my point of view 10% of DC voice coil variance are a absolute no go for a quality driver manufacturer.
Let me make some fun.

Customer: you told me that the right and left speaker are the same?

High end Seller: yes sure

Customer: But one is 1m and the other 1,10m ?

High end seller: oh, thats tolarence has no influence in what you hear.

Costumer: But i hear a difference?

High end seller: you use old copper cables, thats the problem, get this new silver cables and they will sound the same.

;)
Good hypotheses! Or explanations for various error sources.:)

The measurement itself I have carried out with my cheap $15 Euro multimeter. That could be the cause of the error if not the fact that I tested with that multimeter on other speaker elements, se #37. They measure as they should, so the multimeter works as it should.:)

At the moment I'm not home but when I get back I'll see if the measurement via REW gives anything more.
I'll probably have to disassemble the XD125 driver that shows errors. Maybe if it goes well with what MAB suggests:"replacement diaphragms for $20 euros"

The measurements made on the XD125 are run on several occasions. They have the same temperature and so on. They are both in the living room when I have measured with the multimeter. Every time the same result.

Nop clear, the manufacturer says DC resistance 5.7 Ohm, so that should be it. They use a decimal so it should be within that margin of error.
The Peerless tc9 that I measured, seen in #37 is EVEN more accurate with the specs. Peerless state the DC resistance at: 6.32 Ohms. Even less margin of error there. Peerless apparently think their tc9 measures so accurately :) :
Screenshot_2022-11-17_133554.jpg


So if the manufacturer states something, it should be what you get. In my case bought used so I can't complain about that. Let it be a lesson for beginners who read this thread: When buying used speaker elements, study the manufacturer's stated DC resistance and then measure the speaker elements BEFORE the purchase.:)
I actually don't remember how it was with mine. I think the seller just said they measure well. Then I thought no more about it.
Measuring well, that can mean pretty much anything. ;)

Edit:
There is of course the possibility that it was I who burned the driver.
 
Last edited:

Philbo King

Addicted to Fun and Learning
Joined
May 30, 2022
Messages
670
Likes
873
What do you think, or do you have to say about this?

I have measured, and measured again with my (cheap) multimeter. One driver shows 5.6 Ohm and the other driver 6 Ohm. Measurements taken directly on the contacts of the compression drivers. Okay, there's some solder left, but here's what it looks like and sorry for the blurry picture, not easy to measure and take a picture at the same time. :) :
View attachment 243700


This is what the manufacturer states:

View attachment 243699
Edit:
Changed title of the thread, see attached picture.:)
Frankly, the variation in the contact resistance of the ohmeter test leads could easily account for the difference. For ultimate accuracy you need to use a 4-lead ohmeter with the leads soldered to the device under test. Sort of that, start by touching the leads together and have the meter use the 'zero' function to automatically subtract that value from the reading.
 
Top Bottom