Recently I always see myself reaching for the treble knob, so here's my question: can I make my speakers (Triangle Celius 202) sound a bit brighter by changing the value or upgrading the caps on the crossover?
I've read full websites on "the sound" of different caps, especially pricey ones, is this true or just sales talk?
I've read full websites on "the sound" of different caps, especially pricey ones, is this true or just sales talk?
Not really. Assuming that the tweeter has a somewhat standard high pass circuit, changing cap values is only going to impact the lower end of the tweeter's response, and I'll leave it up to others to comment on capacitor type, since I have yet to find any scientifically substantiated information that can provide support for the "audibility" of capacitor types.
Depending on the crossover design, it may be possible to brighten the tweeter if the design utilizes any type of an l-pad or resistive elements to lower the tweeter output to match the midrange. This isn't uncommon in smaller, less efficient speaker designs. Sometimes you can get away with altering those resistor values to raise the output of the entire tweeter, or depending on the design, sometimes changing one resistor can provide a nice upward tilt to the output, but again, this is really application dependant and would likely require modelling the tweeter filter to be able to see what those individual component changes would do.
Depending on the crossover design, it may be possible to brighten the tweeter if the design utilizes any type of an l-pad or resistive elements to lower the tweeter output to match the midrange. This isn't uncommon in smaller, less efficient speaker designs. Sometimes you can get away with altering those resistor values to raise the output of the entire tweeter, or depending on the design, sometimes changing one resistor can provide a nice upward tilt to the output, but again, this is really application dependant and would likely require modelling the tweeter filter to be able to see what those individual component changes would do.
If the original caps are bipolar electrolytic types (the schematic tells they are film but that may be different in real life) and if they're old using film caps will often cause the tweeter to sound brighter than before. It is a simple test to do. You can indulge in buying exotic caps but MKS/MKT are quite good in comparison to bipolar electrolytic caps. MKP or similar are slightly better but they are larger. I would use the same or higher voltage ratings and types that fit exactly as stuff vibrates in speakers. Example: if the original ones are 27 mm and 100V rated I would search for a film cap that is also 27 mm and if this is a 250V type it is OK. All this if quality is the goal. If speed/lack of time/cheap is the goal then don't bother.
* If these speakers are built in 2005 and if they have bipolar caps then replacement seems a logical step for 2 reasons.
* If these speakers are built in 2005 and if they have bipolar caps then replacement seems a logical step for 2 reasons.
Last edited:
Then replacing the 15 µF for MKP will do the trick. Maybe then things can be a little too bright, it sometimes is like that although I assume it will be just right in this case. A 47 Ohm 2W resistor parallel to the tweeter would cure that. Since these don't cost much I would order good quality (= industrial) 2 x 15 µF MKP 5% tolerance and 2 x 47 Ohm 2W metal film resistors "just in case".
As always details are important (please note the askers never seem to give enough information). So please mention brand and series of the non polar caps as used in your speakers. Also carefully inspect them and, if possible, measure them. Regardless of purchase price of these speakers some manufacturers only use the cheapest of cheap.
Good quality non polar caps exist but I for sure have replaced a lot of crappy ones.
As always details are important (please note the askers never seem to give enough information). So please mention brand and series of the non polar caps as used in your speakers. Also carefully inspect them and, if possible, measure them. Regardless of purchase price of these speakers some manufacturers only use the cheapest of cheap.
Good quality non polar caps exist but I for sure have replaced a lot of crappy ones.
Last edited:
Another option to consider, assuming the 15uF cap is still in good shape, is solder a bypass cap across the 4.7uF and 15uF cap ..even something as small as a 0,01uF cap can give a lift to the high frequencies. The choice of cap can influence the perceived sound - eg a silver foil cap will most likely render the brightest sound, but they are generally more expensive.
Well, caps should be Solen or SCR and lythic maybe rebranded at triangle name. I am surprised the Triangle speaker doesn't sound brigth enough ?!
What is your source, have you tried other sources ?
Brigthness is often around 6000 hz.
Indeed bridging with a wire the resistor may inform before further tries: 05R, or 1R but MOX resistor instead of a classic white cement.
You can also try a 4.7 uF from SCR but in Tin foil, or try also before in tin foil a 0.1 uF bypass on the first cap then the second one, then both: experiment to have an idea.
If a SCR aluminium MKP for the 4.7 uF, i.e. same as Solen capacitor, they are known to be neutral, a little smooth. There are better capacitors while in the treble they are often a nice catch and cost nothing, there are more detailed caps but more expensive. A MKT/MKC will sound a little darker in details with a mater background on good reccordings... but for testing purpose on capacitor value they are cheap while the Scr/Solen MKPs are almost as cheap.
If more involved and tooled :
If the second cap is a lythic this is not by chance : designer needed more resistance (ESR) but indeed capacitance value had certainly drifted with a little more capacitance and it can be enough to change it after 25 years. However you don't know exactly how much capacitance value was needed as the bipolar lythic have 5% to 20% variation according the manufacterer or sorting out sometimes on a bench. To have an idea you can measure it to have an idea as the fresh one will need less capacitance -tries and error can be needed to find the value - You can use MH Audio crossover simulator to have an idea of what values were chosen if you know the constructor cut off of this 18 db Buut. filter but it will give you the text book one as you don't know if the filter is assymetric or not (the MKP with a precise value may been chosen for that behavior as MKPs don't drift : you must measure it to know if exactly 4.7 uF.
good luck
What is your source, have you tried other sources ?
Brigthness is often around 6000 hz.
Indeed bridging with a wire the resistor may inform before further tries: 05R, or 1R but MOX resistor instead of a classic white cement.
You can also try a 4.7 uF from SCR but in Tin foil, or try also before in tin foil a 0.1 uF bypass on the first cap then the second one, then both: experiment to have an idea.
If a SCR aluminium MKP for the 4.7 uF, i.e. same as Solen capacitor, they are known to be neutral, a little smooth. There are better capacitors while in the treble they are often a nice catch and cost nothing, there are more detailed caps but more expensive. A MKT/MKC will sound a little darker in details with a mater background on good reccordings... but for testing purpose on capacitor value they are cheap while the Scr/Solen MKPs are almost as cheap.
If more involved and tooled :
If the second cap is a lythic this is not by chance : designer needed more resistance (ESR) but indeed capacitance value had certainly drifted with a little more capacitance and it can be enough to change it after 25 years. However you don't know exactly how much capacitance value was needed as the bipolar lythic have 5% to 20% variation according the manufacterer or sorting out sometimes on a bench. To have an idea you can measure it to have an idea as the fresh one will need less capacitance -tries and error can be needed to find the value - You can use MH Audio crossover simulator to have an idea of what values were chosen if you know the constructor cut off of this 18 db Buut. filter but it will give you the text book one as you don't know if the filter is assymetric or not (the MKP with a precise value may been chosen for that behavior as MKPs don't drift : you must measure it to know if exactly 4.7 uF.
good luck
Last edited:
Are the tweeters working? Check this first.
Solens are gritty, use them across woofers and never on tweeters.
Bypasses don't always work, but 0.1uF is minimum for addition. You can get cheap NOS Russian silver micas in that value, and they do clean up or brighten a smidge.
If removing the resistor is not enough, swap the np cap for a Dynamicap, SoniCap, or MultiCap; in that order most to least bright. These have a brighter signature than some others.
Wolf
Solens are gritty, use them across woofers and never on tweeters.
Bypasses don't always work, but 0.1uF is minimum for addition. You can get cheap NOS Russian silver micas in that value, and they do clean up or brighten a smidge.
If removing the resistor is not enough, swap the np cap for a Dynamicap, SoniCap, or MultiCap; in that order most to least bright. These have a brighter signature than some others.
Wolf
....if he is living in US but if in Europe there should be better choices as Scr tin foil, Jantzen, Audyn Cap or an Mundorf Evo oil.
Re 0.1uF - You might be right technically, but my ears tell me otherwise
I use mundorf (supreme or evo) or a silver foil for higher treble or an audyn copper max if I want the lower treble enhancing.
You may also be right about them not always working ..I can only report on my limited experiences.
Last edited:
Iggy, those brands and models are not what i consider brighter tonality. Audyn is fairly neutral in general. Jantzen if anything has massive soundstage in the upper models, and neutral below. A little bright maybe on the Standard Zs. EVO Oil are dull. SCR/Solen should not be used on tweeters.
Thetall, 0.01uF F3 is above the human range of hearing. It cant affect lower treble. Ive tried 0.01 in more than one instance, and heard nothing. A 0.1 addition even across a resistor for lift can be acknowledged as present.
Wolf
Thetall, 0.01uF F3 is above the human range of hearing. It cant affect lower treble. Ive tried 0.01 in more than one instance, and heard nothing. A 0.1 addition even across a resistor for lift can be acknowledged as present.
Wolf
I agree its above the range of hearing but somehow it influences the overall sound ...at least in my tests. The most obvious difference I heard was when I used a Duelund JDM pure silver 0.01uF. But as you said, I can imagine the effect will vary per system and may not be apparent in many cases.
It happens to have the same model, so I posted the schematic. They are BRIGHT. I sold the triangle tweeters, because even paded down, they were hard. I've tried sb26stcn and scanspeak r3003/602010 in both cases without any changes (except padding) and they both sounded faaaaar better.
By the way, the midrange on this speaker is awesome. The midrange driver is sweet, but those guys at triangle should do a little better with their designs. No padding for a 91 db midrange and only 1 ohm for a high sensitivity tweeter. Those woofers are anyware near matching the sensitivity of mid/tweeter.
It is a newly acquired set of speakers that I got. To be honest I wanted to hear the difference between a good floorstander versus some diy speakers that I own. I think that everyone here in the diyaudio knows, but I have to say it again. The diy speakers are miles ahead of the commersial speakers let alone the triangles (they were 5 star speaker at many hifi magazines)
The most possible scenario for me is to keep the triangles because of they look really nice and they are slim enough to place them, and try to make them sound better.
I plan to keep the original crossover (probably with some changes for the tweeter) with better parts, I will get rid the thermistors, pad the midrange a bit and I will use the scanspeak r3004/602010 for tweeter duties.
So if you guys see anything that could be better on the crossover design please let me know.
By the way, the midrange on this speaker is awesome. The midrange driver is sweet, but those guys at triangle should do a little better with their designs. No padding for a 91 db midrange and only 1 ohm for a high sensitivity tweeter. Those woofers are anyware near matching the sensitivity of mid/tweeter.
It is a newly acquired set of speakers that I got. To be honest I wanted to hear the difference between a good floorstander versus some diy speakers that I own. I think that everyone here in the diyaudio knows, but I have to say it again. The diy speakers are miles ahead of the commersial speakers let alone the triangles (they were 5 star speaker at many hifi magazines)
The most possible scenario for me is to keep the triangles because of they look really nice and they are slim enough to place them, and try to make them sound better.
I plan to keep the original crossover (probably with some changes for the tweeter) with better parts, I will get rid the thermistors, pad the midrange a bit and I will use the scanspeak r3004/602010 for tweeter duties.
So if you guys see anything that could be better on the crossover design please let me know.