2.3 x-over question

[Audioquest]

I am thinking about redoing the crossovers, and after reading that great thread on the 1.2's here is an idea I have that I'd like to bounce off the knowledgable people here.
Seeing as the more parts in a system is not always, but often detrimental, I was considering making a new PCB and eliminating some parts.

Seeing as I am told the 750 pf caps aren't really needed anyway, how about taking this a step further and eliminating a few more parts, but changing values so as not to alter the net effect. In one of the tweeter circuits, the resistors should be the reciprocal of the 2 together.

Comments would be appreciated.....thanks!

Maybe I am totally out to lunch....or at least out for a coffee and sandwich break!

[Joe08867]

Sorry this is above my pay grade.

[Face]

Remove only C4 and the polyswitch, do not make any other alterations. The other 750pf does not appear to be a bypass cap.

[Audioquest]

Remove only C4 and the polyswitch, do not make any other alterations. The other 750pf does not appear to be a bypass cap.

Thanks, other people have told me both 750's don't do anything. So what would happen if I changed the layout, has anyone tried anything else?

[gimpod]

I think if you really study the schematic vs the original board layout I think you'll come to the same conclusion that I have, They DON'T match and there is something wrong here. I've found the same type of differences with not only the SRS 2.3 non-TL but also the 1.2 and 1.2TL. I'm going to be calling Polk in the mourning to see if I can get some answers.

Now I may be wrong here but I've been looking at the schematics vs the original board layout for almost a week and I just can't see how they can match up. Maybe I'm missing something but I know something doesn't jive.

Pic 1 is of a bare board with silkscreen bottom view.
Pic 2 is of the schematic with the error.
Pic 3 is how the schematic should be according to where things are placed on the original board.

[Face]

If your 3rd picture is correct, then yes the 750pf is only a bypass.

[gimpod]

I haven't been able to talk to anybody at Polk about this yet, I'll try next week.

Pic 1 is a top view of an SRS 2.3 (Non TL) crossover and if i'm seeing it right then pic 3 from my earlier post is correct.

More food for thought regarding the SRS 1.2TL crossovers, Pic 2 is the bottom view of an 1.2TL crossover take a close look at how the 12uF cap, 22.5 Ohm resistor, 750pF cap and the jumper between points H and E are connected , it doesn't match the schematic in Pic 3.

Pic 4 is the matching schematic. So could someone tell me what's the point to the 22.5 Ohm resistor and 750pF cap. Because it seems to me that the signal coming from C3 would take the easiest path which would be the jumper at H & E. Now if the 22.5 Ohm resistor & 750pF cap were actually connected to the output of the 12uF cap (C5) that would make since, but it doesn't so once again what's the point, what's it do, why is it there.

Has anybody else noticed this, figured this out or am I on drugs ?

[Audioquest]

Pic 4 is the matching schematic. So could someone tell me what's the point to the 22.5 Ohm resistor and 750pF cap. Because it seems to me that the signal coming from C3 would take the easiest path which would be the jumper at H & E. Now if the 22.5 Ohm resistor & 750pF cap were actually connected to the output of the 12uF cap (C5) that would make since, but it doesn't so once again what's the point, what's it do, why is it there.

Has anybody else noticed this, figured this out or am I on drugs ?

This was exactly why I posed the question of this post. Mine is the 2.3 but it looked to a novice like me, the same situation would occur as you've just pointed out. The 3 caps of the 2.3 are in series, thus just add the capacitances, and the resistors are parallel so take the reciprocal of them. Hence I was asking opinions on why not eliminate the parts.

[gimpod]

This was exactly why I posed the question of this post. Mine is the 2.3 but it looked to a novice like me, the same situation would occur as you've just pointed out. The 3 caps of the 2.3 are in series, thus just add the capacitances, and the resistors are parallel so take the reciprocal of them. Hence I was asking opinions on why not eliminate the parts.

If I were you I would go by the board and NOT the schematic. I don't think it's right. Just my $0.02

[Audioquest]

If I were you I would go by the board and NOT the schematic. I don't think it's right. Just my $0.02

You may well be right. I haven't pulled it out yet, in preparation I was just fielding some questions.

[gimpod]

You may well be right. I haven't pulled it out yet, in preparation I was just fielding some questions.

Always a good thing to do. No harm in asking questions

[FTGV]

So could someone tell me what's the point to the 22.5 Ohm resistor and 750pF cap. Because it seems to me that the signal coming from C3 would take the easiest path which would be the jumper at H & E. Now if the 22.5 Ohm resistor & 750pF cap were actually connected to the output of the 12uF cap (C5) that would make since, but it doesn't so once again what's the point, what's it do, why is it there.

If it actually is connected as in pic4 the 22.5 R and 750p are effectively out of the circuit.Connected as in pic 3 the resistor and 12uf form a contour network that shapes the response of T3.Maybe Polk felt that the shaping was,nt necessary so bypassed it on later production?

The 3 caps of the 2.3 are in series, thus just add the capacitances,

The value of capacitors only sum when in parallel,in series the formula is C= C1x C2/C1+C2.Also when used in a crossover their value can beconsidered independantly depending on which leg of the circuit they are in even though they appear to be in series.

[Audioquest]

The value of capacitors only sum when in parallel,in series the formula is C= C1x C2/C1+C2.Also when used in a crossover their value can beconsidered independantly depending on which leg of the circuit they are in even though they appear to be in series.

Thanks, that was feedback I was looking for

[Audioquest]

The value of capacitors only sum when in parallel,in series the formula is C= C1x C2/C1+C2.Also when used in a crossover their value can beconsidered independantly depending on which leg of the circuit they are in even though they appear to be in series.

Sorry, too much coffee obviously on the Cap math. When does that rule apply that they can be considered independent in some situations?

[FTGV]

When does that rule apply that they can be considered independent in some situations?

Such as when preceeding and following a shunt inductor in a hi pass filter or when paralelled to a resistor in a contour network.

[DarqueKnight]

If I were you I would go by the board and NOT the schematic. I don't think it's right. Just my $0.02

I haven't been able to talk to anybody at Polk about this yet, I'll try next week.

Pic 1 is a top view of an SRS 2.3 (Non TL) crossover and if i'm seeing it right then pic 3 from my earlier post is correct.

More food for thought regarding the SRS 1.2TL crossovers, Pic 2 is the bottom view of an 1.2TL crossover take a close look at how the 12uF cap, 22.5 Ohm resistor, 750pF cap and the jumper between points H and E are connected , it doesn't match the schematic in Pic 3.

Pic 4 is the matching schematic. So could someone tell me what's the point to the 22.5 Ohm resistor and 750pF cap. Because it seems to me that the signal coming from C3 would take the easiest path which would be the jumper at H & E. Now if the 22.5 Ohm resistor & 750pF cap were actually connected to the output of the 12uF cap (C5) that would make since, but it doesn't so once again what's the point, what's it do, why is it there.

Has anybody else noticed this, figured this out or am I on drugs ?

GP,

I received your PM, but I am tied up with holiday festivities and it will be Monday before I can look at this in detail.

I will say that my 1989 version 1.2TL board matched the schematic, with the exception of the order of some series inductors and resistors in the tweeter branches which we discussed previously.

I can say for certain that the 22.5 ohm resistor and 750 pf cap are not connected as shown in your figure 4 in post 7. Sometimes it is easy to get "lost" when comparing the top component side to the bottom trace side. I suggest going back and reverifying the termination points of the components. Holding the board edge perpendicular to the plane of my eyes and matching locations with a pointed instrument on each side of the board helps me to pinpoint component termination points.

I was told that all of the 750 pf silver mica caps were in the circuit to improve the transient response of the mylar film caps and that they were not needed with higher quality film caps. Accordingly, I removed them in my crossover mods.

[gimpod]

I hope these 2 pics help to better demonstrate the point I'm trying to make.

Pic 1 is the bottom side of a 1.2TL crossover (Look familiar DK it should it's yours from this post )
Pic 2 is the bottom view of a striped PCB. Now if someone can tell me how Pin 2 of the 22.5 Ohm resistor is connected to Pin 2 of the 12uF Cap as it is shown in the schematic Pic 3 from my earlier post #7 I'll eat a 1956 Cadillac.

[F1nut]

You know, IIRC, I ran into this when I did Joe's 1.2TL's. I believe I dropped pin 2 of the 12uF into the C5 hole and added a jumper to R2.

[DarqueKnight]

I hope these 2 pics help to better demonstrate the point I'm trying to make.

Pic 1 is the bottom side of a 1.2TL crossover (Look familiar DK it should it's yours from this post )
Pic 2 is the bottom view of a striped PCB. Now if someone can tell me how Pin 2 of the 22.5 Ohm resistor is connected to Pin 2 of the 12uF Cap as it is shown in the schematic Pic 3 from my earlier post #7 I'll eat a 1956 Cadillac.

I traced through the 1.2TL circuit branches using my revision of your crossover back view photograph.

Working backward from Pin 2 of pin header P3 (tweeter connections) and using your pin notation, we encounter:

1. Pin 2 of C5 (12uF),
2. Pin 1 of C5,
3. Juction point D. Point D is also electrically tied to points E, H, and G.
4. There are 5 branches originating at Point D. The branch with the 22.5 ohm is a series combination of a 750pF capacitor, 22.5 ohm resistor, 0.7mH inductor, and 7.5 ohm resistor.

Since points D, E, H and G are all on the same copper trace, this effectively shorts the 750pF cap/22.5 resistor and all that is "seen" on this branch by the signal is the 0.7mH inductor and the 7.5 ohm resistor.

Some Intrique

We know that the 1.2TL crossover board is a modified version of the 1.2 board. It appears, from visually tracing through the circuit traces of the 1.2TL board, that the 22.5 ohm resistor and its associated 750 pF bypass capacitor are shorted out by the common connection of points H and G. This conflicts with the schematic.

If the 22.5 ohm resistor and its associated 750 pF bypass capacitor are supposed to be shorted out, then why include them in the first place? Were there a lot of populated 1.2 boards that needed to be used up or is there some Polk "speaker magic" that explains all of this?

It is inconceivable that Polk would have a tweeter hanging out there with no leveling resistor in the circuit path. If T3's leveling resistor were removed in a later circuit revision, then some adjustment in the component values in the T1, T2, and T4 branches would seem to be required.

The 1.2TL board matches the schematic everywhere else (with the execption of the order of the series R/C and L/R components on the T1- and T4- branches).

Janne mentioned here that one of his 1.2TL's did not have a 22.5 ohm resistor and that the other one had a 22.5 ohm resistor in the 2. ohm resistor's place.

I shorted the 22.1 ohm resistor in my right 1.2TL crossover as shown in figure 1 in order to gauge the sonic difference with that resistor out of the circuit.


Figure 1. The right side 22.1 ohm resistor takes a break.

With T3's leveling resistor out of the circuit, there was a small loss of high frequency detail. There was not a night and day difference between the left and right speaker, but there were missing high freqency details on the right. (e.g.: room ambience, decay, metallic shimmer, female vocals, etc.) . This was apparent when listening to both speakers together and particularly when continuously moving volume control left and right. These high frequency attributes were not missing with the modified stock boards.

Sometime next week, If my dedication to audio holds up, I'll do some frequency sweep analysis of the 1.2TL HF crossover section with the 22.5 ohm resistor in and out, similar to what I did here and here. I'll also dig my stock 1.2TL boards out of storage and examine them.

[gimpod]

DK of course your going to hear a sonic difference by jumping the 22.1 Ohm resistor because by jumping the resistor you are also jumping the 12uF cap seeing how they are wired in parallel, (you jumper one and you also jumper the other). The only way to test this is by removing the 22.1 Ohm resistor. But then again this may be a mute point seeing that you designed your PCB based on the schematic and not on the original PCB.

I think the only way were going to get an answer on whether or not the 22.5 Ohm resistor & 750pF cap needs to be in the circuit is to use a pair of original crossovers and remove the 22.5 Ohm resistor & 750pF cap from one crossover, install the crossovers then do a comparison. If there is a sonic difference then I think there's some Polk voodoo black magic going on here.

But then again maybe it's best just to let sleeping dogs lie.

[Face]

Sounds like a microphone based measurement system is in order. ;)

[DarqueKnight]

DK of course your going to hear a sonic difference by jumping the 22.1 Ohm resistor because by jumping the resistor you are also jumping the 12uF cap seeing how they are wired in parallel, (you jumper one and you also jumper the other). The only way to test this is by removing the 22.1 Ohm resistor.

You're right. I forgot they were wired in parallel. I'll take out the resistor and retest.

I think the only way were going to get an answer on whether or not the 22.5 Ohm resistor & 750pF cap needs to be in the circuit is to use a pair of original crossovers and remove the 22.5 Ohm resistor & 750pF cap from one crossover, install the crossovers then do a comparison. If there is a sonic difference then I think there's some Polk voodoo black magic going on here.

I know that you have contacted Polk CS for an answer. It takes the engineering department a while to respond to these types of requests, sometimes up to a couple weeks or more depending on their workload.

But then again maybe it's best just to let sleeping dogs lie.

I think we should pursue this until we get a definitive answer from Polk.

Sounds like a microphone based measurement system is in order. ;)

Are you volunteering?

[DarqueKnight]

I took the 22.1 ohm resistor out of the right speaker's crossover and it sounded like I had thrown a thin blanket over the speaker, with a definite muddiness to the upper bass.

[gimpod]

This doesn't surprise me seeing how you designed your boards on the schematic and not the original boards. Like I said before I think the only way I think were going to get an answer is to try this test on a stock set of crossovers.

[Joe08867]

Yep, Still above my pay grade.

DK, those are nice boards you got there. Very clean.

[DarqueKnight]

This doesn't surprise me seeing how you designed your boards on the schematic and not the original boards. Like I said before I think the only way I think were going to get an answer is to try this test on a stock set of crossovers.

I sent a very detailed email to Matthew Polk this afternoon regarding this issue.

[gimpod]

I sent a very detailed email to Matthew Polk this afternoon regarding this issue.

Kool, I agree we need to get an answer on this schematic vs circuit board question and not just for the 1.2TL's also the 2.3 and 1.2 Non-TL.