I thought I saw a post about the coil over heating. Maybe I made it up? But I have noticed a slight discoloration around the tweeter in question. It's extremely hard to see with the eye, so I had to adjust the saturation levels in the pic. The second pic is without the adjustments and pretty much how it looks in real life.
The face plate is aluminum. Could the tweeter get hot enough to discolor it?
The face plate is aluminum. Could the tweeter get hot enough to discolor it?
Attachments
Last edited:
Hmm, when did testing drivers at < .25"/6.35 mm from the dust cap [DC] fall out of favor? Not to mention isolating them from all but a signal source?
Yes, there could be a mild hot spot in the VC windings that could cause both the discoloration due to shellac/whatever is sealing the windings and reduced efficiency, so best to pull it out to view/test. While you're at it, might as well test the XOs also since the shelving is so uniform broadband.
Can't answer the poll without sufficient info, but IME the smoke does shift it a bit towards replacement.
GM
Yes, there could be a mild hot spot in the VC windings that could cause both the discoloration due to shellac/whatever is sealing the windings and reduced efficiency, so best to pull it out to view/test. While you're at it, might as well test the XOs also since the shelving is so uniform broadband.
Can't answer the poll without sufficient info, but IME the smoke does shift it a bit towards replacement.
GM
Last edited:
looking at the distortion measurements again I think Mainframe's suggestion of checking for crud in the gap and on the voice coil former is worthwhile. I had thought I had fried a couple of Phillips dome mids years ago and actually replaced them with some inferior cone drivers, after a few years I decided to remove the domes to inspect what had happened, and to my surprise they were not burnt, I cleaned them up and put them back together and was surpised to find that they were fine and the distortion was gone.
I also think that Ralf's suggestion of doing the tweeter swap is advisable as that is the only way you will be 100% sure it is the tweeter and not the electronics that is causing the problem.
I've not looked at enough distortion plots to interpret what is going on, but it is interesting to me that 2nd harmonic is similar with both speakers, but 3rd and 5th harmonic are signicantly higher on the problem one. 7th can't tell unless you do a measurement at 96Khz or 192Khz sampling rate.
I always thought higher order odd harmonics were more likely to be the electronics than the speakers, but this is not something I am sure about.
Tony.
I also think that Ralf's suggestion of doing the tweeter swap is advisable as that is the only way you will be 100% sure it is the tweeter and not the electronics that is causing the problem.
I've not looked at enough distortion plots to interpret what is going on, but it is interesting to me that 2nd harmonic is similar with both speakers, but 3rd and 5th harmonic are signicantly higher on the problem one. 7th can't tell unless you do a measurement at 96Khz or 192Khz sampling rate.
I always thought higher order odd harmonics were more likely to be the electronics than the speakers, but this is not something I am sure about.
Tony.
Ah! In my usual limited time available just scanned down and only 'saw' the 0.5 m. Both are nominally the same except for the notch/distortion and only ~1 dB difference @ 10 kHz/ 0.5 m, yet 'crashes' at the XO point, so could be just the XO or a combination, so again, the procedure is to isolate/test components individually.
As for replacement, especially tweeters, should be done as a ~matched pair [at least same batch order].
GM
As for replacement, especially tweeters, should be done as a ~matched pair [at least same batch order].
GM
Nothing in the manual states that these are implicitly DSP based loudspeakers but everything would be easier if they were 
It's difficult to believe that the firmware, or microcontroller software, would end up corrupted in a way to cause the kind of problems that we are seeing.
It's quite clear that there's a problem with the right speakers tweeter. That's where the elevated distortion and abnormal frequency response is coming from. As mainframe above says, this requires extra investigation.
The obvious and easy solution would be to swap tweeters. This would rule out the electronics. I would be surprised if there was something wrong with them but these things do happen. The tweeter will have it's own DSP channel, own DAC channel and associated analogue circuitry and then its own amplifier channel.
We know the DSP is good as it's been reprogrammed, not to mention a wonky bi-quad wouldn't really cause distortion, just a different filter profile. You say these have been stress tested and performed to specification but clearly that cannot be the case. Either the company is lying, which is possible, or they didn't actually tested the speaker itself. They could have simply removed the electronics module, reprogrammed the firmware, then stress tested the module into dummy loads. This would imply that the electronics modules are functioning properly but the drivers are not. Why a company would neglect to run final frequency response and distortion sweeps on the drivers/finished speakers I do not know, but clearly they didn't. If they had done they'd have seen something wasn't right.
As it stands these look like they use Scanspeak D2905/9x tweeters. Either the 93/95 or 97. You're going to have to unscrew the right speakers tweeter and remove it.
The 93/95 both contain ferrofluid but the 97 does not. Mainframe above says somehow he ended up with a 99 with crud in the magnetic gap. The 99 is very similar to the 97 so it's possible something similar has happened. How? I have absolutely no idea and I'd be surprised if the ferrofluid in a scan tweeter would have dried up in this short a time too.
It's difficult to believe that the firmware, or microcontroller software, would end up corrupted in a way to cause the kind of problems that we are seeing.
It's quite clear that there's a problem with the right speakers tweeter. That's where the elevated distortion and abnormal frequency response is coming from. As mainframe above says, this requires extra investigation.
The obvious and easy solution would be to swap tweeters. This would rule out the electronics. I would be surprised if there was something wrong with them but these things do happen. The tweeter will have it's own DSP channel, own DAC channel and associated analogue circuitry and then its own amplifier channel.
We know the DSP is good as it's been reprogrammed, not to mention a wonky bi-quad wouldn't really cause distortion, just a different filter profile. You say these have been stress tested and performed to specification but clearly that cannot be the case. Either the company is lying, which is possible, or they didn't actually tested the speaker itself. They could have simply removed the electronics module, reprogrammed the firmware, then stress tested the module into dummy loads. This would imply that the electronics modules are functioning properly but the drivers are not. Why a company would neglect to run final frequency response and distortion sweeps on the drivers/finished speakers I do not know, but clearly they didn't. If they had done they'd have seen something wasn't right.
As it stands these look like they use Scanspeak D2905/9x tweeters. Either the 93/95 or 97. You're going to have to unscrew the right speakers tweeter and remove it.
The 93/95 both contain ferrofluid but the 97 does not. Mainframe above says somehow he ended up with a 99 with crud in the magnetic gap. The 99 is very similar to the 97 so it's possible something similar has happened. How? I have absolutely no idea and I'd be surprised if the ferrofluid in a scan tweeter would have dried up in this short a time too.
This is correct Tony but speakers are very good at producing higher orders of distortion if they aren't working properly!
It's possible that the voice coil is rubbing ever so slightly. Or perhaps a transient melted a portion of the coil fusing some of the windings together. Then we'd presumably get an asymmetric magnetic field, less magnetic field strength and the potential for rubbing as the melted winding insulation could have bubbled up.
Something relatively easy you could try jaminjames is to remove the electronics module from both loudspeakers. Disconnect both of the tweeters and measure the DC resistance of the tweeters coils. If one is significantly lower than the other then it's a good bet one tweeter is partially fried. I suppose the other possibility is that the tinsel leads going into the defective tweeter has worn so thin that its resistance has shot up causing distortion, like a faulty connection, and reduced level.
I guess I didn't clarify. I removed the amps and sent them in for service. They were tested in an MM35 cabinet at Barefoot service. It didn't cost much to ship the amps and have them serviced (they also did some updating to wiring harnesses and stuff they do on all older amps). Barefoot is a world class company, and they've been super easy to work with. Just wanted to through that out there.
Interesting problem. Since these are top quality speakers, with good warranty support, I would talk to the manufacturers before taking them apart too much.
Swapping drivers around is dull, but a good plan. See if the problem follows the driver.
But tweeters are simple things to take apart if you are careful with them.
And Scan tweeters are easy enough to fix if something is rubbing or scraping. Tolerances in the voicecoil gap are incredibly tight, but voicecoils are replaceable except with those pointy ring radiators IIRC:
ScanSpeak Voice Coil D2905/9300, D2905/9500, D2905/9700
Questions of ferrofluid complicate things very slightly, but nothing that you can't fix.
Might be worth pressing in the mid to feel for rubbing too. But nothing too invasive without the blessing of the manufacturer at this level. Good companies like to keep customers happy and their reputation good.
Swapping drivers around is dull, but a good plan. See if the problem follows the driver.
But tweeters are simple things to take apart if you are careful with them.
And Scan tweeters are easy enough to fix if something is rubbing or scraping. Tolerances in the voicecoil gap are incredibly tight, but voicecoils are replaceable except with those pointy ring radiators IIRC:
ScanSpeak Voice Coil D2905/9300, D2905/9500, D2905/9700
Questions of ferrofluid complicate things very slightly, but nothing that you can't fix.
Might be worth pressing in the mid to feel for rubbing too. But nothing too invasive without the blessing of the manufacturer at this level. Good companies like to keep customers happy and their reputation good.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.