Welcome, Guest. Please Login or Register
Decware Audio Forums
03/28/24 at 17:05:16 




Pages: 1
Send Topic Print
changing capacitor values on ERR (Read 22563 times)
tom collins
Senior Member
***




Posts: 51
changing capacitor values on ERR
03/25/14 at 16:19:15
 
Guys:  Before I bother Steve, I thought I would see if you all have some suggestions.
Frankly, I love everything about the ERRs but one thing.  On some vocalists, Cat Stevens and now Shelby Lynne, the range of their voices is congetsted and almost cupped-hands like.  Does anyone have any experience in having changed the values of the Caps that are the crossover on this speaker?

Thanks.

Tom
Back to top
 
 
  IP Logged
ZYGI
Seasoned Member
****


HR-ONE

Posts: 757
Re: changing capacitor values on ERR
Reply #1 - 03/27/14 at 01:18:52
 
Tom,

Are you saying congested or might it sound more nasal like?

Zygi
Back to top
 
 

All Decware amps and sources, Turning Point Audio speakers
  IP Logged
tom collins
Senior Member
***




Posts: 51
Re: changing capacitor values on ERR
Reply #2 - 03/27/14 at 22:26:08
 
deeper, throatier, recessed is the best i can describe it.
Back to top
 
 
  IP Logged
will
Seasoned Member
****




Posts: 2914
Re: changing capacitor values on ERR
Reply #3 - 03/30/14 at 02:05:25
 
Tom,

I don't have the ERR but have played with HR-1s. I ended up with minor cap value changes from 3.3 to 3.4 using a Mundorf Supreme 3.3 with a K40Y-9 .1uf wired parallel, taking the range of the tweeter a little lower; and added a .1uf K40 bypass cap from the hot speaker wire terminal in, and bypassing the resistor and main cap so it directly connects the amp to the tweeter, heightening the high highs a bit.

I also really liked the character of Mundorf M-Resist Supreme resistor. Since it was so smooth and solid, it felt less bright, so I compensated using a 2.7 resistor rather than the 3.0 wirewound I had been using.


I tried a number of different caps, and particularly for the bypasses, NOS PIO Russian K40Y-9 really excelled. As an example, some auricaps of the same value sounded unnatural and hard...electronic...whereas the PIOs are smooth and musical.

With the then stock Clarity SA caps, I probably would not have gone with the bypasses as the Clarities were brighter feeling per value than the Mundorfs, but with this Mundorf, the smooth musicality of the K40s bring a balance to the sound that is very real sounding, with more subtle detail, smoothly bringing out textures and ambiance and all very well.

Between the caps and resistor, I made the speakers a bit brighter, but they don't feel bright to me because of the balance and synergy this arrangement gives. The highs are more detailed, without a sense of being overly bright. Beautiful. In retrospect, I think what was happening as I explored (over some months) was that I was optimizing the detail character as much as the quantity, and the deeper I got, the more fascinating the tuning subtleties became.

In my system/room this made the speaker much better... to me anyway.

Maybe you already tried this, but before caps, I was wondering about resistors. It might be interesting to just run a straight copper wire instead of the resister to get a sense of what fully increasing the tweeter's volume does related to this issue....maybe a less resistant resistor would help.
Back to top
 
 

All Modified: Balanced Transformer-DIY Strip/Shunyata Defender-RevolutionMacMini/Amarra-KTE Singxer/Gustardx20pro-ZBIT-CSP3-OldChen 300B/845, Torii IV, Simple Wave 300B-HR1/SVS Micro3000sub-Pi PCs-DIY PCs, ICs, USB, I2S, Speaker-SR and aluminum w ball bearing feet
  IP Logged
Lon
Seasoned Member
****


"Love without
guts is
worthless!"
Philip K. Dick

Posts: 23305
Re: changing capacitor values on ERR
Reply #4 - 03/30/14 at 02:13:17
 
I missed this thread first time around. I agree with Will, try a resistor change. Or I also wonder. .. how old are your tubes (if you're using a tube amp). That sort of recessed and rough vocal sound I have heard when my power tubes had sort of gone past their prime. . . . That was what first came to my mind. I have ERRs that I now have in my parents' system where I do most of my listening. I haven't played around with capacitors, Bob set mine up with me in mind, knowing my experience with the three other sets of speakers he built me before the ERRs, and I think he got that part just right for me. But I have used a number of resistor values and a lower value may also be a fix, especially if your power tubes are in nice shape.
Back to top
 
 

HR-1,ZTPRE,ZBIT,ZROCK2,SEWE300B,CSP3-25mod,Taboo MkIV;Rega RP3 all GrooveTracer mods;PSAudio:PST+DSD,P15,NPC,PowerBases,AC-12 pwr cbls,Reference spkrcbls;Mapleshade SamsonV3;VooDoo:Cremona+Amati interconnects, IsoPods; headphones:Sennheiser HD800S,ZMF Ori,Oppo PM1
  IP Logged
tom collins
Senior Member
***




Posts: 51
Re: changing capacitor values on ERR
Reply #5 - 03/31/14 at 23:34:00
 
thanks for your well reasoned replies.  i will consider this.  the tubes are very fresh and i have 3 sets with all 3 displaying the same quality.

tom
Back to top
 
 
  IP Logged
Lon
Seasoned Member
****


"Love without
guts is
worthless!"
Philip K. Dick

Posts: 23305
Re: changing capacitor values on ERR
Reply #6 - 04/01/14 at 00:32:26
 
Good, it's not the tubes then. Playing with resistors is a good start.
Back to top
 
 

HR-1,ZTPRE,ZBIT,ZROCK2,SEWE300B,CSP3-25mod,Taboo MkIV;Rega RP3 all GrooveTracer mods;PSAudio:PST+DSD,P15,NPC,PowerBases,AC-12 pwr cbls,Reference spkrcbls;Mapleshade SamsonV3;VooDoo:Cremona+Amati interconnects, IsoPods; headphones:Sennheiser HD800S,ZMF Ori,Oppo PM1
  IP Logged
tom collins
Senior Member
***




Posts: 51
Re: changing capacitor values on ERR
Reply #7 - 04/04/14 at 19:03:45
 
incidently, i tried something else with these.  i wondered if the lowness of the mid/woof could be partly responsible since they are so far below ear level.  i brought in some standard  bricks and put them under the speaker.  the stage is definitely taller. i think but am not yet sure about the other problem.  i recently changed rooms and sit much closer, so now it is a greater angle for the sound to reach my ears when seated. i think the bass may be a little more defined as well.
Back to top
 
 
  IP Logged
Matthew (was ML)
Verified Member
**




Posts: 25
Re: changing capacitor values on ERR
Reply #8 - 07/27/14 at 22:53:51
 
How to execute Will's recommendation?  What's the flow of energy?  I'm a novice.  Where do the By-passes go?

T = Tweeter
F = Full range driver
Back to top
 

ERR_Connections_2.jpg

Matthew
  IP Logged
will
Seasoned Member
****




Posts: 2914
Re: changing capacitor values on ERR
Reply #9 - 07/28/14 at 19:37:02
 
ML,

I don't have that setup exactly, but I think your R and C connections are all inline on the positive side of the tweeter. These terminals were done to give us easy access to adjust the tweeter volume and frequency range. With the labeling on the photo it would seem 1 would be the beginning and 2 the end of each R and C, so I think it would go like this. (Might want to PM ZYGI to verify).

The path from the + speaker terminal would be an internal wire connecting it to R1. Then R1 to R2 connect by the resistor you choose, lowering the volume of the tweeter relative to that value. Then another internal wire would connect R2 to C1. C1 to C2 is your cap choice, determining the range from the top frequency into the mids the cap allows to get to the tweeter. After C2 the path would go internal to the tweeter. The negative would run from the minus speaker terminal internally to both the radial and tweeter. Since there is only one radial (mid/bass) driver, there would seem no need for a crossover, so the positive and negative probably run straight to the Radial making its volume and range a given, leaving tweeter adjustments to shift the balance for correct phase, and to our tastes and system/room.

To liven up the top a bit, I would try the "bypass" two ways. One would be to connect the .1 cap wired together with the Mundorf (C-1 to C-2). The parallel connection would add the two cap values, your caps together now 3.4 rather than 3.3.

It has been a while since I read up on this, but there is more to it sound-wise than raising the value a little. Something like this: The low value (.1) bypass cap passes the highest frequency information easily since that is all it does, whereas the Mundorf is working with relatively a greater  and more complex frequency range. So theoretically, if the bypass is a good cap, and its inherent qualities are a little different from the Mundorf, its addition can reveal the top with added complexity and clarity.

This is useful for many as a means to improve the tweeter sound less expensively. You can use a less expensive, decent cap for the main job, and add a small value, high quality bypass cap to make the whole sound better. It will be interesting in your case to hear what you hear since your initial caps are very highly thought of, but I suspect the russian PIO bypasses will change it for the better.

Another way to explore would be to bridge across R-1 to C-2 with the .1 cap, literally bypassing the whole tweeter mediating setup. This would pass the very high information without the resistor or the main cap, and in your case, with the 10 ohm resistor, may allow too much of the highs all else being good, but why not try. Or maybe a .047 there could add sparkle without being too obvious.

My resistor is a big Mundorf M-Resist supreme

http://mundorf.com/en/?category=hifi&menu=resistors_fuses&content=mresist_suprem...

and only 2.7. So bypassing the resistor and main cap has less effect. I am using both methods on mine and really like it, using two- .1 caps.

These resistors with my speakers, system and tastes were well worth the money to me but hard to say for others. As an example of costs partsconnexion has an 8.2 ohm and a 10 ohm listed for 16.52.

I wonder.... have you experimented with inexpensive 9 ohm, 8 ohm, or 7 ohm resistors? If you need just a bit more tweeter shimmer and upper detail, a fine line shift could mean a lot.

My K40Y-9 .1uf caps are higher voltage than those you PM'ed about. One looks the same length as those you saw but is 1000v. The other is shorter and is 630v. Both are .1uf. I see no reason that a 200v would not work fine in a speaker. I think I got the higher voltage ones thinking the heavier construction for higher voltage could be good, but I have no idea if this is true. Also this would allow using them in an amp if I ever wanted to.

Lastly, be sure your caps have long enough wires. I got some once that were pretty short.

Hope this covers it.

Let us know what you find.

Will
Back to top
 
 

All Modified: Balanced Transformer-DIY Strip/Shunyata Defender-RevolutionMacMini/Amarra-KTE Singxer/Gustardx20pro-ZBIT-CSP3-OldChen 300B/845, Torii IV, Simple Wave 300B-HR1/SVS Micro3000sub-Pi PCs-DIY PCs, ICs, USB, I2S, Speaker-SR and aluminum w ball bearing feet
  IP Logged
Matthew (was ML)
Verified Member
**




Posts: 25
Re: changing capacitor values on ERR
Reply #10 - 07/28/14 at 20:09:21
 
Thanks.  I've tried 6 ohms, 8 ohms, 10 ohms and 15 ohms.  I love the extra sparkle on the lower ones but the sound overall looses body so I lowered the tweet to 10 ohms to give more body.  I'm in an apartment and wanting to listen at less than 80 db (neighbors).  I think the large room requires some extra bass and upward firing requires more comparative mids so I think tapping into more mid range forward from the tweet might be great. Also I think 9 ohms may be ideal in my room. But increasing the high sparkle might do it too.  So I'll try the bypass first. Thanks.
Back to top
 
 

Matthew
  IP Logged
Matthew (was ML)
Verified Member
**




Posts: 25
Re: changing capacitor values on ERR
Reply #11 - 08/03/14 at 21:15:33
 
But using the Russian you recommended (value: ,1) in parallel with the Mundorfs silver and oils  (value: 3,3) adds a liveliness and more live sound, which is what I was looking for.  A tad brighter or more extended, not sure.  But nice.  And, as I always do when I interact with this site, I learned something new. Thank you my hifi brother.
Back to top
 

ERR_with_2_caps.jpg

Matthew
  IP Logged
will
Seasoned Member
****




Posts: 2914
Re: changing capacitor values on ERR
Reply #12 - 08/03/14 at 21:26:21
 
ML. That sounds good, and you are welcome. I have found that a little bit can go a long way also, at least considering it in the context of fine-tuned listening and refinement. The last few percent have become kind of huge for me. This "bypass" thing is a really amazing. I am glad it worked for what you were after, even with very good caps to begin with. But then, I find the russians quite refined too.
Back to top
 
 

All Modified: Balanced Transformer-DIY Strip/Shunyata Defender-RevolutionMacMini/Amarra-KTE Singxer/Gustardx20pro-ZBIT-CSP3-OldChen 300B/845, Torii IV, Simple Wave 300B-HR1/SVS Micro3000sub-Pi PCs-DIY PCs, ICs, USB, I2S, Speaker-SR and aluminum w ball bearing feet
  IP Logged
Matthew (was ML)
Verified Member
**




Posts: 25
Re: changing capacitor values on ERR
Reply #13 - 08/05/14 at 04:45:48
 
I spoke with Bob "Zygi."  What a nice man.  

He suggested to try bi-passing from Red + or from #2 to #5 (see photo above of the electricity flow) with a .47 or .68 cap instead of .1.   .1 is very small he says.

Also, my 3,3 cap u is nowhere near the capacity of the tweeter.  Even though 3 is recommended, on this speaker, the ERR tweeter can handle up to about a 12 u or so capacitor before getting near the lower mid-range limit, but at some point cancellation issues with the driver will occur.

Anyone know where cancellation issues show up with the driver?

Zygi also recommended the mundorf M-Resistors.

Seperately, I have an interesting dip in volume around 2000-3000 hz according to my meter.  Somehow, my room is eating that frequency.  Interesting.  Any ideas?
Back to top
 
 

Matthew
  IP Logged
will
Seasoned Member
****




Posts: 2914
Re: changing capacitor values on ERR
Reply #14 - 08/05/14 at 19:15:47
 
Matthew,

I believe you said earlier, that bypassing with the .1 cap from 2 to 5 was too much and that you preferred it parallel from 4 to 5. Also, I did not mention this earlier, but the bypass info I have read said that roughly 10% is a safe max for not messing with phase too much. With a recommended 3 uf cap, and you using 3.3, you were already at 10%, but since going a little past 10% with the HR-1 was fine, I guessed another .1 would be OK for you too. And .1 does a lot  for my sound, while .22,  and .33 were good, but finally ended up being a little too much for me. But that was testing in my particular system and room.

And the ERR is a very different sort of speaker even than the HR-1, with the adjustable tweeter and radial driver, where most speakers have fixed front drivers. And since Zygi is not concerned about a higher value, going further with bigger bypasses could be quite interesting.

As I think about this, the questions would be: Since 2 to 5 is the same as 1 to 5, crossing with the .47 or .68, from 2-5, skipping the current cap and the resistors, would add highs and go deeper into the highs than the .1, but if I recall correctly, you found the .1 in this position overwhelmed what you wanted in the sense of low end balance. Alternately, if you increase the parallel crossover (4-5) by .47 or .68 it would go deeper into the mids, but the questions...would that be too bright for you after the .1 gave you a satisfying sound.

That said, what ZYGI said, the .1 cap only works up high, or in the case of your current arrangement, clarifying the high-highs along with being parallel, bringing the frequency range of your Mundorf down only a little bit  to 3.4. So I am guessing Zygi's thought may be that by having a wider high frequency range in the blend than the .1 provides would be brighter on top, but also a wider frequency range of the highs, and this wider range may sound more balanced and better.

Did Zygi say what the frequency range of your tweeter is, or about what frequency the 3.4 cap crosses it at? If it is the tweeter I think it is http://www.swanspeaker.com/products/products.aspx?cid=9&sid=52&pid=160 like my MG944 and HR-1, then the current version has a recommended crossover at 4500 Hz or above, presumably because of a deep frequency slope-off starting at about 4000, and down about 20dB by 2000.

If it would work without overload, using caps big enough to get down to 2K, trying to address your 2-3K dip area, may help a little even with this steep slope-off, especially if your current caps are already getting you down to 4K. But if your caps are going only to say 6K, a larger cap will crank the 4-6K range much more than the 2-3K range since the tweeter slopes so much below 4K. I think this is right anyway.

Looking for solutions....since your 10 ohm resistor is close to where you want it, an 8.2 ohm M-resist Supreme instead of 10 would be some risk, but you may really like this once you get used to it. The Mundorf is so smooth and solid sounding I remember that these resistors made a lesser value necessary in my system. The Wirewound 3 ohm was brighter feeling than the 2.7 ohm Mundorf. I am guessing the 8.2 Mundorf might bring your tweeter resolution up without sounding much (if any) brighter, but I can't say for sure.

If your room is canceling 2-3K in a notable way, this may be the primary problem in the sense of vibrancy. This is important midrange information for a sense of articulation and especially in the 3K range, there is information that implies presence.

This in mind, speaker placement could be interesting at this point of serious listening. Have you played with the speaker placement and/or tweeter angle lately? Even a little forward and back, and/or a little toe is discernable in my room. From very little up to a foot in any direction could really shift your room reflections...enhancements or cancelations. There may be a place where the balance sounds better to you.

Same with toe and/or tweeter angle.....

Another experiment might be: Since Zygi said up to 12 uf could be OK with the tweeter, if you don't have any other caps around, how about trying 6.6?...If you can set your system for mono...pull the speaker wires from the amp on one channel, and take the Mundorf from that speaker and put it in parallel on the other speaker making it a 6.6 value, and taking it much deeper into the mids. Not an ideal test as it would be only one speaker, but it may give you a sense of the possibilities.

Lastly, I have messed around with a number of favored, but relatively inexpensive caps, and if you want to try Zygi's suggestion of a .47 or .68, Obbligato Gold Premium and Mundorf Supreme were the realist, most natural sounding of the decent but relatively inexpensive caps I tried.
Back to top
 
 

All Modified: Balanced Transformer-DIY Strip/Shunyata Defender-RevolutionMacMini/Amarra-KTE Singxer/Gustardx20pro-ZBIT-CSP3-OldChen 300B/845, Torii IV, Simple Wave 300B-HR1/SVS Micro3000sub-Pi PCs-DIY PCs, ICs, USB, I2S, Speaker-SR and aluminum w ball bearing feet
  IP Logged
Matthew (was ML)
Verified Member
**




Posts: 25
Re: changing capacitor values on ERR
Reply #15 - 08/12/14 at 06:12:46
 
I'm back to 8 ohms R.  
Tried 2x 3,3 cap on each side. Bit deeper.  Nice.

Can't find the 2000 hz issue.  Maybe it is the diver in my ERRs.  I plugged in an old driver from my car and it plays 2000-3000 hz strong and no room issues or interaction with the other ERR.

Tried a .68 bypass (2-5).  It was was super live sounding, cool, but so too bright, harsh, loud in high end that it was overbearing.  

Learning about crossovers.  Found a calculator: http://ccs.exl.info/calc_cr.html

The tweet alone has a nominal impedance of 5 ohm.  If I add a 8  ohm resist should I enter 8 ohms or 5+8=3 ohms into this calculator?

At 8 ohms, if I use 3, it means I can go up to 11,7 u on the cap.  Maybe this why Zygi said I can go to almost 12?  If I use 8 ohms in the calculator, for the resistor alone, then I can only go to 4,4 u to cross at 4500 where the tweet starts to roll off.
Back to top
 
 

Matthew
  IP Logged
will
Seasoned Member
****




Posts: 2914
Re: changing capacitor values on ERR
Reply #16 - 08/12/14 at 07:55:03
 
Sounds like you are getting places toward getting the sound you want.

Best I can figure out you would use the 5 ohm of the tweeter for the crossover calculation since this is what you are putting the cap on to crossover. I think the resistor you use on the way to the caps are before the crossover and function more like a volume control for the crossed tweeter rather than being part of the equation.

When I used the calculator you found, with 5 ohm @ 4500, I got 7.066...5 ohm @ 3000 = 10.6, pretty close to Zygi's estimate. If you tested the two 3.3s and one .68 that would be about 4200, and with two .68s about 4000 right? Could be interesting.

Back to top
 
 

All Modified: Balanced Transformer-DIY Strip/Shunyata Defender-RevolutionMacMini/Amarra-KTE Singxer/Gustardx20pro-ZBIT-CSP3-OldChen 300B/845, Torii IV, Simple Wave 300B-HR1/SVS Micro3000sub-Pi PCs-DIY PCs, ICs, USB, I2S, Speaker-SR and aluminum w ball bearing feet
  IP Logged
Pages: 1
Send Topic Print