Resistors can be used in crossover circuits to attenuate a tweeter which is louder than the woofer.
Resistors can also be placed in series or parallel with capacitors and/or inductors to form impedance equalisation networks.
The PSB crossover is obviously more elaborate than an "off-the-shelf" crossover.
An elaborate crossover can make a loudspeaker system sit up and beg, as well as talk, by a suitable choice of components.
Resistors can also be placed in series or parallel with capacitors and/or inductors to form impedance equalisation networks.
The PSB crossover is obviously more elaborate than an "off-the-shelf" crossover.
An elaborate crossover can make a loudspeaker system sit up and beg, as well as talk, by a suitable choice of components.
Unfortunately, PSB says they do not have a schematic for that crossover. The customer service agent says that not even their techs have it. I guess that's not too surprising given how old those speakers are.
I'm not going to be able to get out to my cabin for at least 2 weeks so for now I guess I'll have to go with my gut which tells me you're correct in your hunch, Galu. Here's what I base it on. I would appreciate it if anyone could provide feedback on whether or not this reasoning is sound.
Take a look at the PSB Image 2Bs impedance curve between 5Khz and 20Khz
As you can see it goes from around 7 ohms to about 8 ohms over that range.
Now look at the impedance curve of the 8 ohm tweeter
An externally hosted image should be here but it was not working when we last tested it.
Look at where the impedance is between 5KHz and 20K.Hz It starts somewhere around 8 ohms range at 5KHz and at 20KHz it's about 11 ohms. So just the impedance of this 8 ohm tweeter on it's own is always about 1 - 3 ohms higher than the impedance of the PSB 2B crossover and tweeter combined at any frequency in that 5 - 20 KHz range.
Now look at the impedance curve of the 4-ohm tweeter.
An externally hosted image should be here but it was not working when we last tested it.
It starts at about 3.5 - 4 ohms at 5KHz and at 20 Khz it's about 5 ohms so it's always about 3 ohms lower than that of the entire speaker the whole way through that range (as Galu pointed out earlier).
I know the tweeter used by PSB in the image 2B was not the exact same as one of these two but it was the same general type. Probably the biggest difference was that it had an aluminum polypropylene dome and these two Vifa tweeters are fabric (silk) domed. The Vifas are reported to be a little more sensitive as well.
If you look at the impedance curve of the 8-ohm Vifa, it stays pretty close to 8-ohms over a lot of its range so it's pretty hard to imagine that PSBs tweeter would've had much lower impedance than this Vifa 8-ohm if it were an 8-ohm tweeter. If we accept that as being true then logic would dictate that PSB's tweeter couldn't have been an 8-ohm tweeter because its impedance alone would be higher than what was reported by the crossover and tweeter in the Image 2B speaker from 5KHz - 20KHz (useful) range.
In looking at the impedance curve for the 4-ohm Vifa, I think you can also assume that if PSB's tweeter was a 4-ohm, it's impedance curve wouldn't likely be a whole lot different than the Vifa 4-ohm's from 5 KHz - 20 KHz. So unless PSB made, or had some other company make them, a custom non-standard impedance tweeter for this and several other of their speakers that used this tweeter (my guess is this is unlikely) logic states that tweeter had to have been a 4-ohm impedance tweeter and cannot have been an 8-ohm. Is this correct/valid reasoning?
Based on this line of reasoning, my guess is that the 4-ohm BC25SC08-04 tweeter is the only one of these two to try as a replacement. If it turns out to be a little hot or bright or whatever you want to call it, you can always tone it down a little with a L-pad if you need to. There will be nothing you could do to dial in the 8-ohm tweeters response the other way if it happens to come in a little dull or dark. Correct?
Last edited:
Using a 4 ohm tweeter with a crossover designed for an 8 ohm tweeter would double its crossover frequency.
This would leave a gap in the response of the speaker system, the audibility of which you would have to judge. Dips are less objectionable than peaks and at least you can be assured that the 4 ohm tweeter would suffer no damage.
Obviously an L-pad can't help restore the crossover point, it can only attenuate the replacement tweeter if happens to be louder than the original.
Without a crossover schematic to inform the experts, and given the urgency of your situation, I can't see you have any option other than to go with your reasoning. Your ears will tell you how valid that reasoning is!
Do let us know how you get on.
Ok, now I think we may have enough information to figure out the schematic on the PSB Image 2B crossover and figure out what the impedance was of the original tweeter.
Attached are photos of the topside and underside of the XO PCB.
When looking at the underside photo, the speaker (+) terminal is on the left. Yellow wire is tweeter (+), blue is tweeter (-). On the woofer, I'm not quite sure which is (+) on the woofer side but it looks to me the iron core inductor jumps from the positive terminal to the black wire so maybe black is (+) on the woofer?
When looking at the topside photo, the cap lowest down and to the right is 8.0 uF. The next one up and sort of in the middle of the board is the 1.5 uF. The other cap at the top is 12 Ohms. I don't know the inductance ratings of any of the inductors.
Right off the get-go, it appears to me that the 3.3 ohm resistor is in series with the tweeter (+). Is that correct?
Lojzek, is this enough information for you to suss out the XO schematic and figure out the most likely impedance of the original tweeter given the response curves I provided earlier?
Attached are photos of the topside and underside of the XO PCB.
When looking at the underside photo, the speaker (+) terminal is on the left. Yellow wire is tweeter (+), blue is tweeter (-). On the woofer, I'm not quite sure which is (+) on the woofer side but it looks to me the iron core inductor jumps from the positive terminal to the black wire so maybe black is (+) on the woofer?
When looking at the topside photo, the cap lowest down and to the right is 8.0 uF. The next one up and sort of in the middle of the board is the 1.5 uF. The other cap at the top is 12 Ohms. I don't know the inductance ratings of any of the inductors.
Right off the get-go, it appears to me that the 3.3 ohm resistor is in series with the tweeter (+). Is that correct?
Lojzek, is this enough information for you to suss out the XO schematic and figure out the most likely impedance of the original tweeter given the response curves I provided earlier?
Attachments
Last edited:
I think I've been able to figure out what goes to what on the XO PCB. I've provided my interpretation in the attached pdf document. If someone with more expertise than I could please check it and input the schematic into a XO design program to see if the design works and what the impedance of the original tweeter most likely is given the impedance curve for the speaker unit, it would be greatly appreciated. If you could put up the schematics for both parts of the crossover it would also be greatly appreciated.
Also, if someone could give an explanation of the XO design, I'd much appreciate it. I'd like to understand how certain aspects of it work and why they did things that way. (e.g. why run the 1.5 uF and 3.3 ohm resistor in parallel to the tweeter? Why run inductors I1 and I2 in series?)
TIA
Also, if someone could give an explanation of the XO design, I'd much appreciate it. I'd like to understand how certain aspects of it work and why they did things that way. (e.g. why run the 1.5 uF and 3.3 ohm resistor in parallel to the tweeter? Why run inductors I1 and I2 in series?)
TIA
Last edited:
Here are some response curves for the PSB Image 2B speaker. The crossover frequency is 2.5 KHz. The upper curve is on-axis, the middle curve is 15 degrees off-axis and the lower one is 30.
Frequency response, 20Hz - 20kHz (measured @ 2m, plotted @ 1m) Loudspeaker sensitivity (averaged 300Hz - 3kHz, 2.83V/1m)
Frequency response, 20Hz - 20kHz (measured @ 2m, plotted @ 1m) Loudspeaker sensitivity (averaged 300Hz - 3kHz, 2.83V/1m)
Last edited:
Thanks Lojzek. With this information are you able to mock this up in your crossover design program and figure out what inductance of the original OEM tweeter must've been to achieve the impedance curve provided several posts up?
If we assume that the original tweeter and the tweeter I want to replace it with are the same impedance (4 ohms) but the replacement tweeter is a few dB more sensitive, where in the tweeter circuit should I place a resistor to tone it down a little bit? Should I increase the size of the 3.3 ohm resistor that's already in the circuit or should I put a second resistor just before the 8 uF cap or put it somewhere else?
If we assume that the original tweeter and the tweeter I want to replace it with are the same impedance (4 ohms) but the replacement tweeter is a few dB more sensitive, where in the tweeter circuit should I place a resistor to tone it down a little bit? Should I increase the size of the 3.3 ohm resistor that's already in the circuit or should I put a second resistor just before the 8 uF cap or put it somewhere else?
I don't know what model the original tweeter is. All I know is that it was a Vifa made for PSB and it was famous for failing at an early age.
The possible replacements (bolt right in) are the Peerless BC25SC08 tweeters. The -04 4 (ohm) and the -08 (8 ohm) versions are available. The 8-ohm is hard to get and the 4-ohm version is very easy to get so that's what I'd rather use and that's what most people use if they don't buy the flawed OEM tweeter from the manufacturer. The BC25SC08-04 is very efficient/sensitive so I suspect one might have to tone it down with another resistor in the circuit to balance it out with the woofer.
Can you please explain why they have the 1.5 uF cap and 3.3 ohm resistor in parallel to each other and in series with the 8 uF cap in the tweeter circuit? In your simulations be aware that I replaced the OEM electrolytic 8.0 uF with a Dayton 5% 8.2 uF audio grade cap if that makes a difference.
Also, I1 is an iron core inductor and I2 is an air core. Can you please tell me why they went with the two Inductors in series instead of just one inductor? I think I know the reason but would appreciate it if you or someone could explain it just in case I'm wrong.
The possible replacements (bolt right in) are the Peerless BC25SC08 tweeters. The -04 4 (ohm) and the -08 (8 ohm) versions are available. The 8-ohm is hard to get and the 4-ohm version is very easy to get so that's what I'd rather use and that's what most people use if they don't buy the flawed OEM tweeter from the manufacturer. The BC25SC08-04 is very efficient/sensitive so I suspect one might have to tone it down with another resistor in the circuit to balance it out with the woofer.
Can you please explain why they have the 1.5 uF cap and 3.3 ohm resistor in parallel to each other and in series with the 8 uF cap in the tweeter circuit? In your simulations be aware that I replaced the OEM electrolytic 8.0 uF with a Dayton 5% 8.2 uF audio grade cap if that makes a difference.
Also, I1 is an iron core inductor and I2 is an air core. Can you please tell me why they went with the two Inductors in series instead of just one inductor? I think I know the reason but would appreciate it if you or someone could explain it just in case I'm wrong.
1.5uF cap was there to lift the HF past 10kHz, and the 3R3 behind 8uF cap is one of the ways to pad tweeter down. 8uF cap with L3 is a 2nd order electrical HP filter.
L1 and L2 form with 12uF a 3rd order electrical LP filter, obviously because designer thought this much was needed to get the response to his liking. I wouldn't be concerned about it.
This is a sim of the HP original filter with L3 estimated, a 4-Ohm version tweeter and the solution could be to place one resistor ( either 4R7, 5R6, 6R8 or similar) before 8uF cap to achieve a total spl of a PSB Image 2B measurement from the review.
L1 and L2 form with 12uF a 3rd order electrical LP filter, obviously because designer thought this much was needed to get the response to his liking. I wouldn't be concerned about it.
This is a sim of the HP original filter with L3 estimated, a 4-Ohm version tweeter and the solution could be to place one resistor ( either 4R7, 5R6, 6R8 or similar) before 8uF cap to achieve a total spl of a PSB Image 2B measurement from the review.
Attachments
Here is a link where you will find the answers you seek with regards to why there is a contour network in the x-over.
Contour Network Designer / Calculator Help
Spend some time there checking out the site. It's a great source to find answers to the questions you have.
Contour Network Designer / Calculator Help
Spend some time there checking out the site. It's a great source to find answers to the questions you have.
Thank you so much Lojzek! And thank you also imjackalope. I didn't know what a contour network was and why they're used before and now I do.
I am going to order those BC25SC08-04 tweeters and some resistors and continue with the upgrade on those speakers. To reduce shipping costs I will order from a Canadian supplier (Solen). I will order some 4R7 resistors plus some that are a little lower and some that are a little higher than 4.7 ohms so I can experiment a little to see what resistance actually sounds best in a real world setting. For those resistors, I plan to order Lynk resistors which are only about a dollar each.
While I'm at it, I think I'll replace those cheap wire wound 3R3s with something better. I can spend a buck to replace them with Lynk resistors or I can spend about $5 to replace them with AchrOhmiC mil spec resistors. I understand the mil spec resistors are made to finer tolerances but other than that, will they give me any audible benefit vs the much cheaper Lynk resistors?
Lojzek, is there anything I could do by way of adding something to or changing something in the tweeter circuit to flatten out the response a bit above 5 - 6 KHz where it starts to ramp up? Is that ramp up due to the effect of the contour network? If so, is there a way to modify that network by changing the C and/or R values in the network to flatten that part of the response curve a little more?
I am going to order those BC25SC08-04 tweeters and some resistors and continue with the upgrade on those speakers. To reduce shipping costs I will order from a Canadian supplier (Solen). I will order some 4R7 resistors plus some that are a little lower and some that are a little higher than 4.7 ohms so I can experiment a little to see what resistance actually sounds best in a real world setting. For those resistors, I plan to order Lynk resistors which are only about a dollar each.
While I'm at it, I think I'll replace those cheap wire wound 3R3s with something better. I can spend a buck to replace them with Lynk resistors or I can spend about $5 to replace them with AchrOhmiC mil spec resistors. I understand the mil spec resistors are made to finer tolerances but other than that, will they give me any audible benefit vs the much cheaper Lynk resistors?
Lojzek, is there anything I could do by way of adding something to or changing something in the tweeter circuit to flatten out the response a bit above 5 - 6 KHz where it starts to ramp up? Is that ramp up due to the effect of the contour network? If so, is there a way to modify that network by changing the C and/or R values in the network to flatten that part of the response curve a little more?
Last edited:
Good to know. I'll just go with the $1 Lynk resistors then. Ill probably get some 3.3 ohms and 4.7s and then some 0.5 and 1.0 ohms so I can experiment in 0.5 ohm increments from 3.3 up to 6.2 ohms by adding the smaller resistors in series with the larger ones. I'm sure I'll hit the sweet spot somewhere in that range.
Thanks again for all your help Lojzek!!
Thanks again for all your help Lojzek!!
I did the tweeter upgrade and it made another large improvement over and above the large improvement already made by replacing the electrolytic caps in the PSB Image 2B XO. I'm very pleased with the difference it made.
@Lozjek, I first started with just swapping the stock PSB/Vifa tweeter for the Peerless BC25SC08-04 (4 ohm) tweeter with no other changes being made to the XO. I did the swap on one speaker and left the other one with the old tweeter and then listened to both carefully to hear the differences.
I could hear right off that the Peerless sounded much more clear, crisp and airy. Much more detail in the high end and clearer mids than the stock tweeter. The stock tweeter sounded a bit muffled in comparison, especially in the female vocals. With the Peerless having much higher sensitivity than the original tweeter, it needed to be "pulled back" a little to keep it in balance with the woofer so I needed to add a resistor ahead of and in series with the tweeter circuit.
On your suggestion (Lozjek), I tried a 4.7 Ohm resistor ahead of the tweeter circuit first. I could've added some smaller resistors to tone the tweeter down a bit more or put in smaller resistors to kick it up a bit with the extra resistors I ordered but I decided to leave it at 4.7 ohm and be done with it. I don't have test equipment or a well trained ear but I thought that the tweeter seemed in good balance with the woofer with 4.7 ohms added. It definitely didn't need to be made any brighter by reducing the resistance ahead of the circuit and I couldn't imagine the sound could be improved by subduing the tweeter further with added resistance so I just left well and good enough alone and went with the 4.7. I still have the other resistors I bought so I can always change it later on if I ever get access to some test equipment to give those speakers a proper response test.
Attached is the final schematic for the upgraded PSB Image 2B crossovers. While I was adding the 4.7 Ohm Lynx resistor to the tweeter circuit I also replaced the original 5W 3.3 Ohm wire wound resistor with a 10W 3.3 ohm Lynx resistor.
Aside from deadening the cabinets (which I intend to do still) the only other simple upgrade that's possible would be to upgrade the Inductors to better ones, especially I1 which is a laminated iron core inductor. Unfortunately, the Inductances of the three inductors are not marked or known and I lack the knowledge and equipment necessary to measure the inductance of the inductors.
A warm heartfelt thanks goes out to @Lozjek and others who contributed to this thread and help me figure out which of the two upgrade tweeter options I should go with. I'm very pleased with the improvements I was able to make to those speakers on a low budget.
@Lozjek, I first started with just swapping the stock PSB/Vifa tweeter for the Peerless BC25SC08-04 (4 ohm) tweeter with no other changes being made to the XO. I did the swap on one speaker and left the other one with the old tweeter and then listened to both carefully to hear the differences.
I could hear right off that the Peerless sounded much more clear, crisp and airy. Much more detail in the high end and clearer mids than the stock tweeter. The stock tweeter sounded a bit muffled in comparison, especially in the female vocals. With the Peerless having much higher sensitivity than the original tweeter, it needed to be "pulled back" a little to keep it in balance with the woofer so I needed to add a resistor ahead of and in series with the tweeter circuit.
On your suggestion (Lozjek), I tried a 4.7 Ohm resistor ahead of the tweeter circuit first. I could've added some smaller resistors to tone the tweeter down a bit more or put in smaller resistors to kick it up a bit with the extra resistors I ordered but I decided to leave it at 4.7 ohm and be done with it. I don't have test equipment or a well trained ear but I thought that the tweeter seemed in good balance with the woofer with 4.7 ohms added. It definitely didn't need to be made any brighter by reducing the resistance ahead of the circuit and I couldn't imagine the sound could be improved by subduing the tweeter further with added resistance so I just left well and good enough alone and went with the 4.7. I still have the other resistors I bought so I can always change it later on if I ever get access to some test equipment to give those speakers a proper response test.
Attached is the final schematic for the upgraded PSB Image 2B crossovers. While I was adding the 4.7 Ohm Lynx resistor to the tweeter circuit I also replaced the original 5W 3.3 Ohm wire wound resistor with a 10W 3.3 ohm Lynx resistor.
Aside from deadening the cabinets (which I intend to do still) the only other simple upgrade that's possible would be to upgrade the Inductors to better ones, especially I1 which is a laminated iron core inductor. Unfortunately, the Inductances of the three inductors are not marked or known and I lack the knowledge and equipment necessary to measure the inductance of the inductors.
A warm heartfelt thanks goes out to @Lozjek and others who contributed to this thread and help me figure out which of the two upgrade tweeter options I should go with. I'm very pleased with the improvements I was able to make to those speakers on a low budget.
Attachments
Last edited:
.png){kind=link}
.png){kind=link}