All,
My first post here. I've sent my CS2.2 tweeters off to Thiel to be rebuilt since I blew them by using incorrect jumper setting on outboard preamp. In the meantime. I'm looking to re-cap the stock crossovers since they are around 20 years old now. The xover appears has several NP electrolytics being bypassed by 1uF Elcap Polystyrenes/Foil. I guess I have at least two options; 1. Just replacing electrolytics with same value OR 2. Upgrading to some nice film/foil and doing away with the bypasses (except for the parallel 100uF's). Any suggestions? I'm leaning towards just changing out the electrolytics in the spirit of keeping the original design. My plan is not having to go back into these for a while. I tried to attach a pic of the mid/tweeter crossover. Many thanks for any suggestions...
My first post here. I've sent my CS2.2 tweeters off to Thiel to be rebuilt since I blew them by using incorrect jumper setting on outboard preamp. In the meantime. I'm looking to re-cap the stock crossovers since they are around 20 years old now. The xover appears has several NP electrolytics being bypassed by 1uF Elcap Polystyrenes/Foil. I guess I have at least two options; 1. Just replacing electrolytics with same value OR 2. Upgrading to some nice film/foil and doing away with the bypasses (except for the parallel 100uF's). Any suggestions? I'm leaning towards just changing out the electrolytics in the spirit of keeping the original design. My plan is not having to go back into these for a while. I tried to attach a pic of the mid/tweeter crossover. Many thanks for any suggestions...
Attachments
Makes sense to replace parts with drifted values far beyond/below specified. Give each NPE part a capacitance test and then decide.
I infer from your post that you vote to replace with originally designed NPE's. Would it not make more sense to proactively replace them all instead of just the "drifters"? I'd rather spend my time listening to music than checking caps in the future. Thanks
Um, personally, the smallest Solen caps are what I would be most interested in replacing.
Using Mundorf SIO or Clarity CMR are my top choices, followed by Clarity CSA and Mundorf MKP.
Also replace any resistors in series with the mid/tweeter with Mills.
The biggest caps are usually shunts. They are the most expensive, and most fussy. You must measure the total ESR of the entire network (NP + Film) and replace with similar cap and ESR or same cap + ESR + R.
My go-to brand for BIG film caps is AXON, which is essentially SCR/Solen, but far cheaper.
Best,
E
Using Mundorf SIO or Clarity CMR are my top choices, followed by Clarity CSA and Mundorf MKP.
Also replace any resistors in series with the mid/tweeter with Mills.
The biggest caps are usually shunts. They are the most expensive, and most fussy. You must measure the total ESR of the entire network (NP + Film) and replace with similar cap and ESR or same cap + ESR + R.
My go-to brand for BIG film caps is AXON, which is essentially SCR/Solen, but far cheaper.
Best,
E
Ok. Really big thanks for the info. Seems like it may be easiest to change out NPs to same value (thinking Mundorf Ecap) then upgrade solens with mundorf sio. I dont see how upgrading shunts are going to do anything for me regarding SQ. Resistors would also be fairly easy to upgrade. When you talk ESR, wouldnt any of these changes affect overall ESR as well? Just how sensitive are these networks to E S R? Seems a resistor with 10% tolerance could make even a bigger difference than the shunts. Im confused.
Thanks.
Thanks.
ESR sensitivity depends on where in the circuit the cap is, what kind, and how old.
Film caps all tend to have relatively similar and very low ESR ( < 0.5 Ohms). Replacing them tends to be completely safe.
Older non-polar electrolytics (NPE)s rather high (1-3 Ohms). Replacing a cap that naturally has 2 ohms of ESR with a new film cap of 0.3 Ohms can make a huge and hard to diagnose difference, ESPECIALLY if it's a shunting cap. Seen it half a dozen times. This makes the overall impedance wonky, which even sturdy amps have trouble with, and suddenly the whole speaker sounds off. That's because the "bad" cap may be in the woofer low pass, but actually affect the tweeter output. Lastly, the ESR in a shunt cap can affect the phase matching between speakers. I often tweak this to get it right in my own designs. 1-2 ohms difference here is the difference between perfectly aligned drivers and not.
OTOH, replacing an NPE cap that is in series, say for instance, on a mid-range driver is far less risky, but you run the risk of increasing the output of the driver and creating an imbalance that way too.
In either case, the "fix" is to measure the ESR before hand and make sure that you compensate for any lost ESR with additional R if needed.
Best,
E
Film caps all tend to have relatively similar and very low ESR ( < 0.5 Ohms). Replacing them tends to be completely safe.
Older non-polar electrolytics (NPE)s rather high (1-3 Ohms). Replacing a cap that naturally has 2 ohms of ESR with a new film cap of 0.3 Ohms can make a huge and hard to diagnose difference, ESPECIALLY if it's a shunting cap. Seen it half a dozen times. This makes the overall impedance wonky, which even sturdy amps have trouble with, and suddenly the whole speaker sounds off. That's because the "bad" cap may be in the woofer low pass, but actually affect the tweeter output. Lastly, the ESR in a shunt cap can affect the phase matching between speakers. I often tweak this to get it right in my own designs. 1-2 ohms difference here is the difference between perfectly aligned drivers and not.
OTOH, replacing an NPE cap that is in series, say for instance, on a mid-range driver is far less risky, but you run the risk of increasing the output of the driver and creating an imbalance that way too.
In either case, the "fix" is to measure the ESR before hand and make sure that you compensate for any lost ESR with additional R if needed.
Best,
E
Last edited:
BTW, I use Dayton DATS V. 2.0, however you can put together an accurate jig using Room EQ Wizard as well:
Impedance Measurement
Impedance Measurement
Given the fact that you'd rather enjoy listening to music than dealing with XO network again, you are absolutely right in your assumption. I am the kind that exchanges parts once they are shot, not before and occasionally I fancy to turn the soldering iron on.
This is great info. I was not aware of the effects of ESR on these networks. Should I expect these kind of differences between the old NPEs vs new ones if I dont try film caps? Or in other words, can I simply replace the electrolytics with new electrolytics without needing to rechecking the ESR?
Pickinjim, age matters.
The only one's "safe" to update IMHO are any in series with a driver, so a mid-range most likely.
The only one's "safe" to update IMHO are any in series with a driver, so a mid-range most likely.
I should also say, it's not a big deal if you get it wrong. I mean, if you replace a series cap on the midrange, and it's now too loud, we add a resistor later. It's not the end of the world. 🙂
Best,
E
Best,
E
In this case where you are getting brand new tweeters, perhaps you should leave everything alone until you've actually heard them? Then once you're used to them again, you could play with the tweeter portion of the xo.
For now, you should measure the electrolytics in circuit and compare L and R just to ensure they are the same. They might still be in 'good enough' shape after all this time. To accurately measure them may require them to be desoldered, but then you might as well replace with new electrolytics.
For now, you should measure the electrolytics in circuit and compare L and R just to ensure they are the same. They might still be in 'good enough' shape after all this time. To accurately measure them may require them to be desoldered, but then you might as well replace with new electrolytics.
Ok. Well, don't think I need to go any further. I'll be speaking with Thiel service about the blown tweeters, I'll ask about the XOs. Too many ways I can mess things up and I just dont have a lot of time to be fooling around. Thanks a lot for the info. Probably saved me a bunch of headaches.
I always enjoy it when I stumble onto something that so nicely blends art and science. Seems like crossover design may just be one of these things. For these speakers, maintenance is best left to the very experienced. I absolutely cherish them and their place in history and I certainly don't want to make things worse. Just want to do what I can to keep them alive another 20 years. Thanks again. I'll keep you posted.
Well Pickinjim,
I have an online design, with all the simulation files built in. You may use it to play with ESR if you'd like. 🙂
A Speaker Maker's Journey: The LM-1 Bookshelf Version
Best,
E
I have an online design, with all the simulation files built in. You may use it to play with ESR if you'd like. 🙂
A Speaker Maker's Journey: The LM-1 Bookshelf Version
Best,
E
Spoke with Thiel. Electrolytics shouldnt be an issue at this point. They are sending a new resistor that needs updating. I guess film caps are still fair game. I'll look into options. Current recommendations are very expensive. Thanks
Jim,
Since you were upgrading I was recommending top price stuff. The Mundorf MKP caps are to my ears far far superior to Solens. Cheap too, around $8-$10 each.
Best,
E
Since you were upgrading I was recommending top price stuff. The Mundorf MKP caps are to my ears far far superior to Solens. Cheap too, around $8-$10 each.
Best,
E
I could probably get by with Supreme EVO aluminum oil if this is a step up. The solen is 2uF. Mundorf has 1.8 or 2.2uF. Should I go with a 1.8 and still get by? Is 2.2 going to pass too much LF signal? Thanks in advance.
Wow, those prices are pretty good! But pay attention to the physical (length, diameter) sizes too! The MKP's are about the same as the Solen's, but the more expensive models are larger. The point is, don't buy a cap you can't fit! 🙂
Also, you only need 100V, but these caps come in 450V and 600V varieties. More V = physically larger AND more expensive, so stick with the lowest possible V.
I'd try to stay as close as you can. Parallel 1uF:
1uF + 1uF = 2uF
or
1.8uF + 0.22uF = 2.02uF
Also, you only need 100V, but these caps come in 450V and 600V varieties. More V = physically larger AND more expensive, so stick with the lowest possible V.
I'd try to stay as close as you can. Parallel 1uF:
1uF + 1uF = 2uF
or
1.8uF + 0.22uF = 2.02uF
ESR measurement of a NPE vs. MKP
I have had an idea to check how NPE ESR may keep up with a poly type so I gave myself a little work to satisfy curiosity. Intertechnik states for glatte folie caps a tangent delta of 0,023 at 1 kHz. One might assume this to be a constant value, and it appears that in this duel the difference is not all that great but evident. A small cap value like this will have greater influence at higher frequencies so the 1 kHz ESR spec is not very much detrimental.
I have had an idea to check how NPE ESR may keep up with a poly type so I gave myself a little work to satisfy curiosity. Intertechnik states for glatte folie caps a tangent delta of 0,023 at 1 kHz. One might assume this to be a constant value, and it appears that in this duel the difference is not all that great but evident. A small cap value like this will have greater influence at higher frequencies so the 1 kHz ESR spec is not very much detrimental.
