SDA SRS 3.1TL - driver phase?

DollarDave

I picked up a pair of 3.1TL's and am checking them out for a crossover upgrade. The drivers seem to be wired weird. After the first two drivers, the wiring harness swaps the black and the white wires...I expected them to be wired in parallel, but, the harness has me thrown off.

My question is - are these supposed to be wired in parallel, or are two of the stereo drivers supposed to be out-of-phase?

I'll toss in another question. The cross-overs have a 2 Ohm , 10%, 25Watt resistor in the dimensional network - the schematics don't show that. Keep it, toss, it, upgrade it to Mills, and if upgrade, what wattage?

gimpod

Dave, I've never seen the wiring in the 3.1tl but from my knowledge of the 4th & 5th generation crossovers they are expecting a nominal impedance of about 8 Ohm's (in reality about 6.5 Ohms) on the stereo side and about 4 Ohms (in reality about 3.5 Ohms) on the dimensional side.

Now on the stereo side the 3.1TL uses 4 MW-6503's (6.54 Ohms each) so Polk wired them in whats called a parallel/series arrangement where 2 sets of 2 drivers each are wired in parallel (giving us 3.27 Ohms per set) and then the 2 sets of drivers are wired together in series (giving us 6.54 Ohms total). They may look out of phase but there not.

Now that 2 Ohm, 10%, 25 Watt resistor has NO business being in that circuit period! To be honest I have no idea why it would be there, maybe to bypass the 18mh sub-bass inductor or make up impedance for the wrong dimensional driver it should be one MW-6511 (3.13 Ohms).

Can you post any pictures of the crossover? This would help figure out whats going on here, What ever it is it doesn't sound right.

lawdogg

I will try to check mine tonight and let you know what I find.

DollarDave

The dimensional driver is a 6511 and the stereo drivers are 6503s

Crossover pics

gimpod

Dave the pictures helped a lot. I've never seen a 3.1TL crossover look like this. It looks liked someone wire that 2 Ohm resistor in series with the 18mh sub-bass inductor (that big coil of wire the board's mounted to) and it should not be there. The black wires that come from the 18mh inductor go to holes P and R on the board. Again that resistor should not be there.

Now I could be wrong but I don't really think I am. Maybe PM Trey (VR3) and see if he'll take a look as he's seen more 3.1TL crossovers than I.

Question: Did they seem a little week in the bass department.

DollarDave

Hard to say about the bass, I got them with the tweeters missing! That's an easy project while I upgrade the crossovers...

I'll rework them as you suggest and see how they sound.

The wiring harness actually matches what you described. The confusing part was how the wire harness feeds the top two first, then comes back to the bottom two. Top two in parallel then the bottom two in parallel from the negative side of the top two.

headrott

Tony is correct. The 18mH sub-bass inductor (wired in series with the 2 ohm resistor) is absolutely not wired correctly. Someone F-ed up on these crossovers. Also, the 2 ohm inductor is incorrect. It should be (I believe) 1.3 ohms if I remember correctly. If/when you replace it, a 10 or 12 watt will be fine. I recommend Duelund cast resistors, personally. That is, if you plan on keeping these a long time. Duelunds are phenominal. Mills are good, but not as resolving as Duelund.

I will need to double check how the drivers are wired as I don't remember for sure.

That said, the drivers are all correct. The dimentional are 6511's and the stereo are 6503's, so you are good there.

DollarDave

headrott wrote: »

Tony is correct. The 18mH sub-bass inductor (wired in series with the 2 ohm resistor) is absolutely not wired correctly. Someone F-ed up on these crossovers. Also, the 2 ohm inductor is incorrect. It should be (I believe) 1.3 ohms if I remember correctly. If/when you replace it, a 10 or 12 watt will be fine...

Are you referring to a different 2 ohm dcr inductor, or the 2 ohm resistor that is in series with the 18mH? If so, which one is 2 ohm?

Or, are you referring to the resistor that is stood up? That is a 1.3 ohm.

headrott

Ooops, yes Dave I am referring to the resistor, not an inductor.

I am referring to both resistors, actually. I was thinking the one standing up is also a 2 ohm (along with the other flat laying resistor). I cannot see the value, but was thinking that you were saying both are 2 ohm.

The resistor standing up (if it is 1.3 ohm) is correct in value (and wiring). The other resistor (which is 2 ohms) should not be there at all.

Lastly, the 18mH inductor should be re-wired to the P and R holes (as Tony said).

Sorry for my mis-wording in the last post. Hopefully it's more clearly stated now.

DollarDave

Thanks all! Very helpful.

Now, if someone has a set of front grills in any condition, PM me.

decato

No need to worry...no one messed with those crossovers. You have an early release of the 3.1tl. You are lucky that your left and right channels match. I purchased a pair a while back, and I had mismatched crossovers. The 2 ohm resistor is paired with a 20-gauge, 18.0 mH inductor. In later models, Polk eliminated the resistor and simply used a 22-gauge, 18.0 mH inductor.

Basically Polk engineers decided to increase the resistance of the inductor. I imagine that the single dimensional driver exhibited high excursions when the speaker was pushed. The 22-gauge inductor has about the same DC resistance as the 20-gauge inductor and 2 ohm resistor in series. The end result sounds somewhat similar, and it's less expensive to use thinner wire and easier to not have to deal with a series resistor.

Hope this helps,
Brian

gimpod

@decato, Now that I've thought it over for a bit and studied the 3.1TL schematics I think your right that these are very early run of 3.1TL's, But I think you came to a wrong conclusion as to why Polk made the change in the inductor. It had more to do with increasing the overall nominal speaker impedance to be more compatible with the amps of the day. Most amps from those day did not like driving speakers with a nominal impedance of less than 8 Ohms. And remember SDA"s can drop as low as 1-2 Ohms at times.

Today's amps don't mind driving low impedance speakers as a mater of fact if he pulls that 2 Ohm resistor out and wires that inductor as suggested it may improve bass response as this is a mod suggested by the man himself in this paper Inductor Upgrades For SDA.pdf. This file along with a bunch of others about SDA Technology used to be here somewhere but apparently got lost after or during the switch to this POS "Hello Kitty" style forum.

decato

gimpod wrote: »

@decato, Now that I've thought it over for a bit and studied the 3.1TL schematics I think your right that these are very early run of 3.1TL's, But I think you came to a wrong conclusion as to why Polk made the change in the inductor.

@gimpod, How so? The 3.1tl has a single dimensional driver. The 1.2tl has four dimensional drivers connected in series/parallel. The overall impedance would be the same in each (sub)system. Furthermore, the early release of the 3.1tl used a 20-gauge 18 mH inductor which definitely has more resistance than the 1.2tl's 18-gauge 16 mH inductor. If we follow with this reasoning, one might argue that Polk engineers should have added a series resistor to the 1.2tl as well.

So, since the 3.1tl with the 20-gauge inductor already has a higher impedance than the 1.2tl, why would an additional resistor be added in series? I believe the single dimensional driver exhibited high excursions. Take the series resistor out and play some fairly loud test tones at 20, 25, and 32 Hz, and compare the movement of the drivers in the stereo line array with the single dimensional driver. By increasing the resistance, the dimensional driver's movements are kept in line with those of the stereo array, allowing each component in the whole system to reach its limit at about the same point. The thing is, the 22-gauge inductor does exactly the same thing, and it's less expensive!

DollarDave

I ordered a 2 ohm mills to put in place of the sand jobber. Any suggestions on a different value?

gimpod

DaveMuell wrote: »

I ordered a 2 ohm mills to put in place of the sand jobber. Any suggestions on a different value?

Dave, First I wouldn't use that mills unless it is at least 25 watts, As far as I know mills doesn't make a 25 watt resistor. If it was me I would try it with out the resistor and see how it sounds. If you have a decent amp I don't think you'll run into any so called high excursions of the dimensional driver. But if you insist you could use 2 4 Ohm 12 watt resistors in parallel for a combined 2 Ohm 24 Watt resistor, Unless you want to use a plain sand resistor.

This next part is probably going to piss somebody off, but so be it.

@decato

Too begin with you obviously didn't read the paper or you just skimmed it. Just in case you can't figure out how to use a .pdf reader I quote the paper here:

Inductor Upgrades For SDA’s

“The thread on the forum has correctly identified most of the concerns. Good
job!

However, I can add these comments. Usually, in a typical low pass crossover, one would not want to swap out an inductor for another one with much different characteristics because it would alter the voicing of the speaker. However, in this case there could be a performance benefit if the amp being used doesn't mind seeing a lower impedance. The inductor I believe they are talking about is part of the "Full Complement Sub-Bass Drive" circuit. That's a term that Sandy coined to describe the system we used to allow the SDA drivers to work in parallel with the stereo drivers in the bass while producing the SDA signal at higher frequencies.

In this generation of the SDA's the stereo drivers were nominally 6.5 ohms and the SDA drivers were nominally 3.5 ohms. The plus terminals of the SDA drivers on each side were connected to the plus amp terminal on that side via their cross-over network. Then, the negative terminals of the SDA drivers on one side were connected to the negative terminals of the SDA drivers on the other side via the interconnect cable. This causes the SDA drivers on the right to produce an R-L signal and the ones on the left to produce L-R. Since bass is pretty much mono in most recordings, if the SDA drivers both get full-range R and L signals they would cancel at low frequencies and the SDA drivers would just sit there acting like badly tuned passive radiators. So, we added an inductor in each speaker between the SDA driver negative terminal and the amp negative terminal on that side. At low frequencies that diverts the signal back to ground instead of through the interconnect to the SDA drivers in the other speaker. As a result, at very low frequencies the Right SDA drivers produce only right channel bass and vice versa, while still producing the R-L and L-R signals at higher frequencies. The transition occurs between about 50Hz and 150Hz.

However, because of the DC resistance of the inductor, the system isn't perfect. I don't recall the DC resistance of these coils but it was high, at least several ohms depending on the model. This means that the SDA drivers will continue to produce some SDA signal even at very low frequencies. Decreasing the DC resistance will definitely improve the bass response of the system both qualitatively and quantitatively. However, before you rush out to buy those Hi-Q replacement inductors be aware of some concerns.

We chose the higher DC air core coils for a couple of good reasons. First, we were always on the edge of acceptability with the impedance of the SDA's. The DCR of these coils kept the minimum impedance high enough for the amps available in those days. Depending on the model, reducing the DC resistance of these coils may take the minimum impedance down to around 2 ohms. If your amp doesn't mind, you shouldn't either. The other concern is saturation of the inductor core. Air core inductors don't saturate. Given the cost of adequate ferrite or laminated core inductors at the time, plus the need for a higher DCR, the air core choice was obvious. So, when switching to a ferrite or laminated core inductor make sure it will handle at least 5 amps without saturation. That's equivalent to 100 watts of low frequency power through the SDA driver.

hope this helps.”

-Matthew Polk

Resistance is not always futile. The DCR of our inductors is accounted for in the crossover design. Changing it / lowering it will affect a different balance than the design intent. We typically use large enough wire that the DCR of the woofer inductors is in tenths of an ohm range. As to switching inductor for reasons of other qualities, there can be benefits if the designer has not done his homework. We’ve used air-core inductors for many designs and steel or iron cores for others. We base our decisions on the perceived use of the product, cost and size. Iron or steel cores increase the value of a given inductor by focusing the magnetic field created by the windings. This has the advantage of reducing the size and number of winds needed to achieve an inductance value. This also means that larger wire or fewer turns can be used to achieve a lower DCR. The concern is that the core (iron, steel laminate, ferrite) will saturate a some maximum field intensity and pushing current above that level into the inductor will make it become non-linear.

The inductor is actually reverting to acting like an air core as it is over-driven – but only for the over-driven portion of the signal. So the signal becomes distorted, bad noises, scratchy, etc.. We have to choose the wire gauge and core material so as to provide head-room for the largest signals (including transients) that we expect the speaker to reproduce linearly.

Air-core inductors do not saturate as more current is passed through them unless the current is so great that the wire begins to heat. It is not necessary, however, for the inductors to have such high limits because there are plenty of other practical limitations on loudspeaker output. Many of these are simply physical, like the maximum possible cone excursion of the drive units. Go beyond this and once again – distortion. Typically we can design our speakers with steel-laminate inductors and easily reach undistorted levels well in excess of 100 dB in the case of appropriately sized units. The down side of air-core inductors is as mentioned earlier that they will require more wire and hence have higher DCR. So more power from the amp will go into heating up this inductor and less into making sound. In this case, resistance is indeed futile.

As an addition to the thoughts I gave earlier on inductor quality you should advise that only laminated steel is acceptable for most any high quality system. Ferrite saturates at current levels much too low. It’s high permeability allows one to make very high inductance values with less wire (so lower DCR) but the current range is poor at best. For all I know this may be the reason that all quality / high current transformers use laminated steel.”

-Stu Lumsden

(The only changes I made to the above quote are to some minor spelling errors and to bold, underline and put in red the point I was trying to make.)

Also to compare 3.1TL's to 1.2TL's electrically would be like comparing apples to oranges. They are 2 different animals plus you forgot or engorged the fact that the dimensional and stereo drivers and inductors are tied together in parallel with one another. I believe Polk made the changes to better voice the speaker and make it more compatible with the mid to low end amps of the day seeing how the 3.1TL's could be considered the low end of the SDA SRS line.

decato

DaveMuell wrote: »

I ordered a 2 ohm mills to put in place of the sand jobber. Any suggestions on a different value?

I too wouldn't recommend putting a 12-watt resistor in series with the woofer. You can either keep what you have, put 2 Mills resistors in parallel, or get the inductors rebuilt with 22-gauge wire.

decato

@gimpod

This next part is probably going to piss somebody off, but so be it.

Thank you for being yourself. I love when people show their true colors. Now it's my turn.

Too begin with you obviously didn't read the paper or you just skimmed it.

I believe you mean "to". "Too" means also. Perhaps you just skimmed the dictionary. I'll supply you with a link: http://dictionary.reference.com/

Just in case you can't figure out how to use a .pdf reader I quote the paper here:

Thank you again, I couldn't possibly have figured that out how to open a PDF with multiple degrees in computer science. However, given all the schematics I posted, scans of articles I shared, and images I generated for you and others on this forum, it's clear that I know my way around a computer. Furthermore, I have already read that post.

Also to compare 3.1TL's to 1.2TL's electrically would be like comparing apples to oranges.

They share more in common than you might think. After all, didn't you make the same replacement circuit board for all the SRSs?

They are 2 different animals plus you forgot or engorged the fact that the dimensional and stereo drivers and inductors are tied together in parallel with one another.

The laws of physics and principles of electrical engineering haven't changed. Series, parallel, and series/parallel still operate the same, whether the model is called 3.1tl or 1.2tl. In case you don't know the difference, read here: http://www.eminence.com/support/wiring-diagrams/

I believe Polk made the changes to better voice the speaker and make it more compatible with the mid to low end amps of the day seeing how the 3.1TL's could be considered the low end of the SDA SRS line.

You had originally stated that someone tampered with the other fellow's crossovers. We now know better on that topic. Despite all of the arguing, you have not given me any evidence as to why the early release of the 3.1tl had a 2 ohm resistor in series with a 20-gauge inductor while the later releases had a 22-gauge inductor. If you can give me detailed information on the answer to precisely this question, please write me (filled with insults, if you wish). Otherwise, don't bother me again, and have a great life.

lawdogg

Sorry I took a long time here are mine, one side is vertical the other horizontal mounting. They are not sequential serials #s, I don't know why.



same one



and other side

gimpod

^^ We'll spank me like a dead monkey that makes 2 now!. Theses must be early Polk crossovers for the 3.1TL. This kinda weird because no where in the schematics for the 3.1TL does Polk make any mention/note that they removed the 2 Ohm 5% 25 Watt resistor the only note the do make on the schematic is and I quote "5/3/90 | A (with a circle)| DEL POLYSWITCH; 22 GA WAS 20GA; 1.3ohms WAS 1ohm | MPB"

Well looks like I've proven wrong on these crossovers. Looks like I've got some up dating to do.

And finely to @decato I owe you a big apology for my post to you, a lot of it was uncalled for, moronic, and beneath me. For this I am truly sorry. This is no means an excuse (because there is none for what I posted) but an explanation. I woke up yesterday angry for lack of sleep and pain then I came with in a hairs breath of ending up on the floor trying to get out of bed (my attendant screwed the pooch) as this is not a good thing because if I had hit the floor I could have spent the whole day in the hospital getting poked, prodded and x-rayed to death and things went downhill from there and you ended up taking the brunt of my anger and for that I am again truly sorry for that. I sincerely hope you accept my apology.

Tony (aka gimpod)

P.S. @decato I'll try and put together a logical answer to your question on the inductor change in the next day or so. The big problem in trying to figure this out is not having the actual stock DCR of the 1.25mh 18 GA stereo inductor or the stock DCR of the 18mh 22 GA and 20 GA Sub-Bass inductors.

headrott

Agreed Tony. There is nothing stating that the 2 ohm 25W resistor was removed and the 18mH inductor was changed from 20 guage to 22 guage in the schematics. Obviously, this doesn't mean it didn't happen (obviously) because we have two examples of it happening now.

That said, can DaveMuell and lawdogg confirm that the 18mH inductors are, in fact 20 guage, also what the listed dates are on your crossovers?

gimpod

I think it would have to be prior to 05/1990. Also I just noticed something the part number on the tag is BE0031-A, where in the SDA Series Driver And Crossover List it's listed as BE0031-B. So there were 2 versions of the 3.1TL's now all we have to do is figure out why, which may never be answered because the guys who designed these speakers are long gone from Polk.

headrott

The prior to 05/90 date would be my guess as well Tony. I noticed the month on Dave's crossover states 04, but cannot see the year.

Do we know for a fact that two different guages (20 guage and 22 guage) are used for the 18mH inductor? My guess is they would be since these two pictured crossovers have that 2 ohm 25 watt resistor. But do we know for a fact that there are two different guages?

gimpod

For a fact on the 2 different gauge inductors, no we don't. I believe there were but until we can get someone to measure the wire diameter (20 GA = 0.03200", 22 GA = 0.02540") and DCR we won't know for sure. Maybe one of the 2 guys has a micrometer and can measure the inductor wire diameter and while there at it if they can measure the DCR her's a link on how to do that Inductor DCR Insanity For SDA's that would help.

decato

gimpod wrote: »

I owe you a big apology for my post to you, a lot of it was uncalled for, moronic, and beneath me. For this I am truly sorry.

No problem. Everyone has a bad day once in a while. Hope you are doing better now.

decato

headrott wrote: »

Do we know for a fact that two different guages (20 guage and 22 guage) are used for the 18mH inductor?

Yes, when I purchased my pair, the left and right channels were mismatched. Take a look at the images I've attached. The 20-gauge inductor is labeled BE0032 A, and the 22-gauge one is labeled BE0033-A. In the second image, the 22-gauge inductor is on the bottom. Notice how much more of the plastic bobbin is visible, as the wire is taking up less space.

lawdogg

Shoot, I'll have to pull it apart again to check the date. Sorry, I should have thought to look. Will update ASAP.

headrott

decato wrote: »

headrott wrote: »

Do we know for a fact that two different guages (20 guage and 22 guage) are used for the 18mH inductor?

Yes, when I purchased my pair, the left and right channels were mismatched. Take a look at the images I've attached. The 20-gauge inductor is labeled BE0032 A, and the 22-gauge one is labeled BE0033-A. In the second image, the 22-gauge inductor is on the bottom. Notice how much more of the plastic bobbin is visible, as the wire is taking up less space.

Very interesting @decato. So, did/does only the 20 guage 18mH inductor crossover and 3.1TL (from 03/90) have the 2 ohm 25W resistor and the 3.1TL with the 22 guage inductor (from 8/90) not have the 2 ohm 25 W resistor (this would be my guess)? It seems rediculous Polk would have paired these 3.1TL's together since they are so different (in a physical sense) yet I realise they are similar but not idential in an electrical sense.

Are your 3.1TL's still like this? Or did you redo the crossovers and boards? If you did, did you alter anything on the circuit board? Or keep them the same? If they are the same, can you hear/feel a difference in the low frequency output between the two 3.1TL's?

I have never seen a pair of 3.1TL's with a 20 guage 18 mH inductor and/or the 2 ohm 25W resistor until this thread. Also, as stated by Tony previously, why is the change from the 20 guage inductor and 2 ohm 25W resistor to the 22 guage inductor and no 2 ohm 25W resistor not mentioned in the schematic?

Thanks decato.

decato

headrott wrote: »

So, did/does only the 20 guage 18mH inductor crossover and 3.1TL (from 03/90) have the 2 ohm 25W resistor and the 3.1TL with the 22 guage inductor (from 8/90) not have the 2 ohm 25 W resistor (this would be my guess)?

Yes, that's right.

Are your 3.1TL's still like this? Or did you redo the crossovers and boards? If you did, did you alter anything on the circuit board? Or keep them the same? If they are the same, can you hear/feel a difference in the low frequency output between the two 3.1TL's?

No, I had the inductor re-spooled with 22-gauge wire to make a matched pair. The circuit board is the same. Furthermore, the internal bracing was loose in one cabinet, and the passive radiators were also mismatched (A and B versions). So, I had some work to do before it was even worth a serious listening. Sorry, I cannot be specific about the differences in sound.

decato

I also remembered that this problem has been floating around the forum for years. Here's one of the earlier posts:

http://forum.polkaudio.com/discussion/comment/1652650

gimpod

decato wrote: »

headrott wrote: »

So, did/does only the 20 guage 18mH inductor crossover and 3.1TL (from 03/90) have the 2 ohm 25W resistor and the 3.1TL with the 22 guage inductor (from 8/90) not have the 2 ohm 25 W resistor (this would be my guess)?

Yes, that's right.

Are your 3.1TL's still like this? Or did you redo the crossovers and boards? If you did, did you alter anything on the circuit board? Or keep them the same? If they are the same, can you hear/feel a difference in the low frequency output between the two 3.1TL's?

No, I had the inductor re-spooled with 22-gauge wire to make a matched pair. The circuit board is the same. Furthermore, the internal bracing was loose in one cabinet, and the passive radiators were also mismatched (A and B versions). So, I had some work to do before it was even worth a serious listening. Sorry, I cannot be specific about the differences in sound.

You don"t happen to recall the part number of the "A" version of passive radiators, do you recall if they were flat lick a sda2 or 1 or were they cone shaped like the be versions. I could be wrong but something is telling me that I ether read or herd that at one time the 3.1TL used an SW-121 or SW-120 passive radiators.

I could be wrong but with the change in the crossover and the passive radiator I now believe it was done to change the voicing of the speaker more than anything else.

Now Matthew Polk did state "However, because of the DC resistance of the inductor, the system isn't perfect. I don't recall the DC resistance of these coils but it was high, at least several ohms depending on the model. This means that the SDA drivers will continue to produce some SDA signal even at very low frequencies. Decreasing the DC resistance will definitely improve the bass response of the system both qualitatively and quantitatively. However, before you rush out to buy those Hi-Q replacement inductors be aware of some concerns.

We chose the higher DC air core coils for a couple of good reasons. First, we were always on the edge of acceptability with the impedance of the SDA's. The DCR of these coils kept the minimum impedance high enough for the amps available in those days. Depending on the model, reducing the DC resistance of these coils may take the minimum impedance down to around 2 ohms. If your amp doesn't mind, you shouldn't either."

Now will lowering the DCR of that inductor in the 3.1TL ("A" or "B" version) have a bad effect on the single MW-6511 I don't know, And I don't think ether one of us will until someone sits down and test it. Were not going to get an answer from Polk because the guys who designed this speaker are long gone from Polk and doing other things with there life's.

To be honest I never knew there were 2 versions of the 3.1TL's, I guess you do learn something everyday. I just wish I could get the DC R of both the 18mh 20 GA and 18mh 22 GA inductors along with the DCR of the 1.25mh 18 GA inductor.

lawdogg

Real quick follow up. I couldn't find the build dates of the crossovers, but the terminal plate had a sticker indicating 5/21/90 date of assembly. (Edit: I totally forgot to measure the gauge of the windings!) I also snapped some pics of one of my PR's since there was some discussion. No, I haven't dynamatted them.