Change in Resistance with Polyswitch Removal

Stew

This may be old news to some but it recently occurred to me that I changed the resistance of the tweeter circuit in my SDA 2B's when I removed the polyswitches. I removed the bypass capacitors from the tweeter circuit some time ago. They sounded more natural but were missing some of the sparkle. Some time later I removed the polyswitches and the sparkle returned – maybe too much. They sound a little too bright now at times.

I started wondering exactly what I had changed and tried to find specifications for the RDE050A and RDE090A polyswitches that Polk used in the original monitors and SDA's. Although I didn't find specifications for these polyswitches, I did learn that, in general, polyswitches have a base resistance and the resistance increases rapidly when exposed to current above a certain level. Then I found a big clue on the resistance value from this schematic for the original model 5b:

http://www.polkaudio.com/forums/attachment.php?attachmentid=18765&d=1155784608

In this schematic, Polk removed the fuse and replaced with a RDE050A polyswitch. They also changed the resistor in the tweeter circuit from 2.7 ohms to 2.0 ohms indicating that the base resistance of the RDE050A is 0.7 ohms.

The RDE090A used in later models appears to have the same resistance based on this schematic for the 1989 version of the SDA 2B's:

http://www.polkaudio.com/forums/attachment.php?attachmentid=26676&d=1186345413

Polk changed the RDE050A to RDE090A but did not change the value of the resistor in the tweeter circuit indicating that the RDE050A and the RDE090A have the same resistance (0.7 ohms). I measured the resistance of the RDE090A's that I removed from my 2B's and they do measure about 0.7 ohms.

I think I remember reading somewhere on this forum that the RD0198-1 tweeters in my SDA 2B-TL's measure 5.6 ohms. If this is right, then removing the polyswitches changes the overall resistance in the tweeter circuit from 5.6+2.7+0.7=9.0 ohms to 5.6+2.7=8.3 ohms, about an 8% decrease in resistance.

This begs the question of what value resistor to use in the tweeter circuit when the polyswitches are removed. In the case of polyswitches removed and bypass capacitors installed, I suspect they would sound considerably bright. In this case, it might be worth increasing the resistors by 0.7 ohms. In the case of both polyswitches and bypass capacitors removed, the best sounding value for the resistor may lie somewhere between the original value (2.7 ohms for my 2B's) and +0.7 ohms (3.4 ohms for my 2B's). Has anyone tried adjusting the resistor value as a result of removing polyswitches?

Stew

P.S. I'm planning to buy a handful of small value resistors to see what value sounds best. Once I decide on a value, I'll replace the 2.7 ohm resistor with something between 2.7 ohms (the original value of the resistor) and 3.4 ohms (the original value of the resistor + polyswitch). The easiest way to do this would be to add resistors at the tweeter terminal. Does it make any difference if I temporarily install them at the tweeter terminal vs. the original location? I know the resistance will be the same either way but does it affect anything else such as crossover frequency?

MillerLiteScott

Very interesting.

I am all ears.

Scott

Face

Someone correct me if I'm wrong.

When using a RD0-198 in 2B's/CRS's and the 5.8uf capacitor is added to the 2.7ohm resistor, that section of the crossover becomes a contour circuit, and adding or subtracting resistance to the 2.7 ohm resistor will only affect a certain frequency(@6khz) opposed to the tweeter's level as a whole.

heiney9

Hey guys,

I notice in many of the Monitor series that use the sl2000 there is 2.0 ohm resistor if they are fused and a 2.7 ohm resistor is they use a polyswitch. I believe the polyswitch does alter the circuit in a small way. With the right cables and equipment I don't doubt a subtle difference can be heard. I'm going to experiment with my (fused) 5B's with a 2.7 and 3.5 ohm just to take a bit of the fowardness away. I have the RD0194-1's and they aren't harsh or grainy it just seems they are a bit "loud" in the upper vocal range. I hope to have the upper vocals blend a little better.

Let us know what you find out.

H9

P.s I installed a 2.0 ohm resistor per what was already on the board as stock.

Stew

Face wrote: »

Someone correct me if I'm wrong.

When using a RD0-198 in 2B's/CRS's and the 5.8uf capacitor is added to the 2.7ohm resistor, that section of the crossover becomes a contour circuit, and adding or subtracting resistance to the 2.7 ohm resistor will only affect a certain frequency(@6khz) opposed to the tweeter's level as a whole.

Face - I know just enough about circuits to be dangerous and am not familiar with contour circuits. The diagram for the TL modification does add a 5.8uF cap to the 2.7 ohm resistor as you stated:

http://www.polkaudio.com/forums/attachment.php?attachmentid=26710&d=1186358580

I was originally planning to increase the 2.7 ohm resistor to a larger value. Based on your comment, I assume that I should add a small value resistor in another location. Does it need to be in the same location as the original polyswitch or can I put it at one of the tweeter terminals?

ben62670

If you put the resistor at the terminal you will alter the XO point.

Stew

Face - I've been thinking about this and I understand what you mean by contour circuit. The resistance of the polyswitch, however, will affect all frequencies going to the tweeter because it is located before the contour circuit (5.8uf cap and 2.7 ohm resistor in parallel). Thanks for pointing this out. I might have really screwed up things if I had changed the 2.7 ohm resistor.

Ben - I was afraid of that; Thanks for confirming. It will be a little more difficult to try different values but I'll try a small resistor where the original polyswitch was.

ben62670

Stew wrote: »

...The resistance of the polyswitch, however, will affect all frequencies going to the tweeter because it is located before the contour circuit ...

I am pretty sure you have a grip one what the resistor does in place of the Poly switch, but just to clarify it. The resistor would just attenuate all the frequencies going to the tweeter not manipulate them. I have just bypassed them, and many others have to with no ill effects. I would try just jumping the Poly switch, and if it is too bright, or forward after giving the XO time to burn in you could add a 1ohm in place of the jumper.
Ben

Stew

Ben - The poly's are currently removed/jumped. When they were in place, the sound was a little too dull and recessed. With them jumped, the sound can be a little bright at times. I upgraded the crossovers a couple of years ago so they should be broken in. I just recently pulled out the polyswitches.

I think the higher frequencies will see about an 11% decrease in resistance when the polyswitches are removed assuming 5.6 ohms for the RD0198-1's and 0.7 ohms for the polyswitches. I'm also assuming that the highest frequencies see the two capacitors in series but don't see the 2.7 ohm resistor. Does this sound right. Maybe I'm splitting hairs but I was thinking of trying a 0.5 ohm Mills MRA-5.

ben62670

If it is to loud on the highs I would use a 1ohm 12 watt. You can stand it on end;) Please keep in mind the values you are looking at are the DCR of the tweeter, and not the whole impedance curve;)
Ben

Stew

Ben - I'm going to take your advice and give it some time before I do anything. Some CD's sound excellent and some sound a little bright. Not harsh or grainy just a little bright. All sound a little more forward than with the polyswitch but I like the more forward sound. Anyway, I'm going to give it some time and listen to a variety of CD's before changing anything. Thanks for the help!

F1nut

ben62670 wrote: »

If you put the resistor at the terminal you will alter the XO point.

I have to disagree. It's common practice to add a resistor at the tweeter (positive lead) to attenuate it. Now, adding a cap like CL does will alter the crossover point. Therefore, I strongly discourage adding a cap.

Stew, try a .5 ohm, 12 watt to start with. I think you'll find anything smaller has little to no effect and anything larger causes too much output loss.

ben62670

F1nut wrote: »

I have to disagree. It's common practice to add a resistor at the tweeter (positive lead) to attenuate it. Now, adding a cap like CL does will alter the crossover point. Therefore, I strongly discourage adding a cap.

Stew, try a .5 ohm, 12 watt to start with. I think you'll find anything smaller has little to no effect and anything larger causes too much output loss.

Adding a resistor at the tweeter will raise the whole impedance curve by the value of the resistor which in turn affects the XO point;) This is one of the main reasons for 2 different RDO's.

HB27

Basically Polyswitches are inconsistent fuse protection if you listen at higher volumes. There is convenience is not having to change the fuse but they add additional resistance. Once heated or tripped the resistance is higher than initial resistance resulting in inconsistant highs. Resistance can also raise in the polyswitches over time or abuse resulting in diminished highs or even failure.
Though a polyswitch is somewhat faster and safer than a fuse it's not idiot proof by any means and complete protection is far from the case.
To me it makes more sense to eleminate the polyswitch, add a resistor in the tweeter + lead to compensate for the initial resistance and/or use for tweeter hf attenuation, and mount a fuse on the outside of the cabinet for convenience and it's limited protection. I usually use a 1/2 amp fuse and find it more than adequate at the clean power levels I normally run. Clean power being the key.
Harry

Stew

F1nut wrote: »

Stew, try a .5 ohm, 12 watt to start with. I think you'll find anything smaller has little to no effect and anything larger causes too much output loss.

F1 - I'm looking for a subtle change so I think this is what I'll try.

Thanks,
Stew

DarqueKnight

stew wrote: »

this may be old news to some but it recently occurred to me that i changed the resistance of the tweeter circuit in my sda 2b's when i removed the polyswitches. I removed the bypass capacitors from the tweeter circuit some time ago. They sounded more natural but were missing some of the sparkle. Some time later i removed the polyswitches and the sparkle returned – maybe too much. They sound a little too bright now at times.

I removed all the bypass caps in my SDA's and I also either removed or shorted the polyswitches in my SDA's. I heard a more natural high frequency presentation, but it was not forward or bright. It was as if a veil had been lifted off the high frequencies. Another analogy would be cleaning a light coating of dust off a window. When the dust is removed, the light coming in from outside does not get brighter in absolute terms, the light just appears brighter because you are just able to see it better.

stew wrote: »

...i started wondering exactly what i had changed and tried to find specifications for the rde050a and rde090a polyswitches that polk used in the original monitors and sda's. Although i didn't find specifications for these polyswitches, i did learn that, in general, polyswitches have a base resistance and the resistance increases rapidly when exposed to current above a certain level.

The polyswitches Polk used in the SDA's and the current replacement polyswitches were made by the Raychem corporation. Doing a search on the part numbers lead me to Raychem's website (www.raychem.tycoelectronics.com) and to the websites of vendors who sell the parts. The old RDE series part numbers were superceded by and are now cross referenced to RUE and RXE series part numbers:

Part number RDE090A was replaced with part number RUE090.

Part number RDE050A was replaced with part number RXE050.

Part number RDE070A was replaced with part number RXE075.

The current polyswitches that are provided by Polk to replace the RDE parts are Raychem part number RXE135.

The RXE135, RXE075, and RXE050 are obsolete, but are still available from parts vendors. They are even sold on eBay. The RUE090 is still in production, but it is being phased out by Raychem.

Data sheets for the RUE090 and RXE series polyswitches are attached below.

The RXE050 (RDE050A) has a resistance range of 0.50 to 0.77 ohm, with a post trip resistance 1.17 ohm.

The RUE090 (RDE090A) has a resistance range of 0.07 to 0.12 ohm, with a post trip resistance 0.22 ohm.

The RUE090 (RDE090A) has a resistance range of 0.25 to 0.40 ohm, with a post trip resistance 0.60 ohm.

The current replacement polyswitch, the RXE135, has a resistance range of 0.12 to 0.19 ohm, with a post trip resistance 0.30 ohm.

I have four unused RXE135's. Their measured resistances were 0.16, 0.16, 0.17, and 017 ohm. I saw that the schematics for the SDA 1C, SDA SRS 2.3TL, and SDA SRS 1.2TL specify the RXE135 polyswitch. However, the polyswitches removed from my SDA 1C's were RDE050A's. I still have one of the polyswitches taken from my 1C's. It measured 0.85 ohm. While this is not a post trip level of resistance, it is well out of spec.

I also have five of the RDE090A polyswitches which came out of my three pairs of 1989 model SDA CRS+'s (I seem to have lost lost, misplaced, or mistakenly thrown away the other polyswitch. The 1989 SDA CRS+ uses the same crossover as the 1989 SDA 2B, yet the 1989 CRS+ schematic specifies the RDE050A and the 1989 2B schematic specifies the RDE090A. Maybe Polk neglected to update the 1989 CRS+ schematic.

The measured values of the resistances for the RDE090A's taken from my CRS+'s were 0.29, 0.22. 0.23, 0.21, 0.23, and 0.42 ohm. These values are in the range for an RDE090A that has been tripped.

stew wrote: »

then i found a big clue on the resistance value from this schematic for the original model 5b:

http://www.polkaudio.com/forums/attachment.php?attachmentid=18765&d=1155784608

in this schematic, polk removed the fuse and replaced with a rde050a polyswitch. They also changed the resistor in the tweeter circuit from 2.7 ohms to 2.0 ohms indicating that the base resistance of the rde050a is 0.7 ohms.

The rde090a used in later models appears to have the same resistance based on this schematic for the 1989 version of the sda 2b's:

http://www.polkaudio.com/forums/attachment.php?attachmentid=26676&d=1186345413

The data sheets lead me to think otherwise.

stew wrote: »

polk changed the rde050a to rde090a but did not change the value of the resistor in the tweeter circuit indicating that the rde050a and the rde090a have the same resistance (0.7 ohms).

No. The data sheets indicate different resistances for the RDE070A and RDE090A. As I discuss later, I don't think a change in resistor value was required.

stew wrote: »

I measured the resistance of the rde090a's that i removed from my 2b's and they do measure about 0.7 ohms.

The specified resistance range for a non-tripped RDE090A/RUE090 is 0.07 to 0.12 ohm, with a post trip resistance of 0.22 ohm. The average pre-trip resistance of the RDE090A/RUE090 is probably approximately 0.10 ohm. I think it is safe to say that your polyswitches have been tripped several times.

stew wrote: »

i think i remember reading somewhere on this forum that the rd0198-1 tweeters in my sda 2b-tl's measure 5.6 ohms. If this is right, then removing the polyswitches changes the overall resistance in the tweeter circuit from 5.6+2.7+0.7=9.0 ohms to 5.6+2.7=8.3 ohms, about an 8% decrease in resistance.

Based on the spec sheet for the RDE090A/RUE090, the DC resistance of the tweeter circuit is more likely 5.6 + 2.7 + 0.1 = 8.4 ohms.

The stock ceramic metallic resistors were 10% tolerance, 5 watt parts. The DC resistance of the tweeters has some tolerance value also, although I do not know what it is. Assuming it is also 10%, the series DC resistance range of the resistor/tweeter combination would be 8.3 ohms +/- 0.83 ohm. Clearly, removing 0.1 ohm of resistance would not be significant. It certainly would not be significant to the point of requiring compensation.

We must also consider that the absolute resistance value of a resistive element is not the only thing that affects the sound. Ideally, a resistor would only attenuate the signal passing through it. Actually, resistors attenuate and, like every other component, introduce some noise and distortion.

While the attenuation properties of the polyswitch are relatively insignificant, the noise characteristics of the devices are considerable and audible...even when they are brand new. After a polyswitch is tripped the first time, the resistance increases from 52% to 83%. The noise from the "burned" polyswitches also greatly increases. [Quantifying the signal damage done by new and tripped polyswitches would be a good project for an interested meter pontiff.:)] Removing the polyswitches removes a tiny bit of attenuation, but removes a significant amount of noise. When noise is lowered, the signal becomes apparently louder although the amplitude (level) of the signal remains the same. It is analogous to cleaning a dirty window.

stew wrote: »

this begs the question of what value resistor to use in the tweeter circuit when the polyswitches are removed. In the case of polyswitches removed and bypass capacitors installed, i suspect they would sound considerably bright. In this case, it might be worth increasing the resistors by 0.7 ohms.

My SDA's did not sound considerably bright with the polyswitches removed and the bypass capacitors still in place. When the bypass capacitors were used with "faster" polypropylene film capacitors, artificially enhanced "air" and "ambience" was introduced on some recordings. Some people like this effect. I do not.

When I consulted with Polk prior to doing my first SDA modification, I was advised to remove or short the polyswitches if and only if I was using quality amplification and if I was not in the habit of overdriving my amplifier. There was no advisement on a requirement for resistance compensation due to the removal of the polyswitch.

We must also consider that four different polyswitches were used in the various SDA models: RDE050A, RDE070A, RDE090A, and RXE135. Each of these polyswitches has different resistance specs. However, the current replacement for the polyswitches in all SDA's is the RXE135.

The forwardness or brightness you are hearing could be due to other components in your audio chain or even the source material you are playing. The increased detail with the polyswitch removed is probably just now revealing sonic characteristics that were previously masked.

stew wrote: »

in the case of both polyswitches and bypass capacitors removed, the best sounding value for the resistor may lie somewhere between the original value (2.7 ohms for my 2b's) and +0.7 ohms (3.4 ohms for my 2b's). Has anyone tried adjusting the resistor value as a result of removing polyswitches?

Again, I do not think this is required. You stated that your speakers:

stew wrote: »

...sound a little too bright now at times.

Usually, a coloration is consistently evident. However, colorations can be more or less apparent depending on the characteristics of signal running through the component. Again, I would ask you to evaluate your audio chain from source material to source components to amps and the interconnects and speaker cables in between. In absolute terms, I don't think making significant changes in the resistance of the tweeter circuit is the answer. In subjective terms, if you prefer the sound with more resistance added to the tweeter circuit, there is nothing wrong with that.

Season to taste.

Stew

DK - Thanks for taking the time to provide such a thorough answer!! Maybe I'll try replacing my POS cables instead.

Thanks,
Stew

Stew

DarqueKnight wrote: »

The RXE050 (RDE050A) has a resistance range of 0.50 to 0.77 ohm, with a post trip resistance 1.17 ohm.

The 1989 SDA CRS+ uses the same crossover as the 1989 SDA 2B, yet the 1989 CRS+ schematic specifies the RDE050A and the 1989 2B schematic specifies the RDE090A.

Interesting - I assumed the polyswitches in my 1989 2B's were RDE090A's as that is what the schematic calls for. I just took a closer look at them and they are labeled 050. They were tripped one or two times with a smaller amp - about 18 years ago. It does sound significantly cleaner with the polyswitches out; Just a little too bright on some CD's.

DarqueKnight

SDA's are sneaky...and they love to deviate from published specs.;)

F1nut

DarqueKnight wrote: »

SDA's are sneaky...and they love to deviate from published specs.;)

Yeah, like those nasty 750pF bypass caps that were in one of my sets of SDA's.....rated at +/- 50%. You read that right, 50%!

DarqueKnight

You talk like that's a bad thing.

Stew

A pair of 12W 0.5 ohm Mills resistors showed up in the mailbox today and I installed them where the polyswitches used to be. The change is subtle but that's all I wanted. I'm listening to Dire Straits BIA right now and it's sounding really good!

P.S. I measured the RDE050A's again and 1 measured about 0.7 ohms and the other measured a little over 1 ohm. I must have tripped one channel but not the other many years ago.

HB27

Stew wrote: »

A pair of 12W 0.5 ohm Mills resistors showed up in the mailbox today and I installed them where the polyswitches used to be. The change is subtle but that's all I wanted. I'm listening to Dire Straits BIA right now and it's sounding really good!

P.S. I measured the RDE050A's again and 1 measured about 0.7 ohms and the other measured a little over 1 ohm. I must have tripped one channel but not the other many years ago.

You can trip or damage a polyswitch just by turning the amp on or off if you have an amp that "pops or makes noise" on start up or shut down. I had an Adcom that did that on a regular basis and lost a tweeter even when the polyswitch tripped.
Harry

george daniel

hb27 wrote: »

you can trip or damage a polyswitch just by turning the amp on or off if you have an amp that "pops or makes noise" on start up or shut down. I had an adcom that did that on a regular basis and lost a tweeter even when the polyswitch tripped.
Harry

ouch



I had a 555 that did the same thing,, fortunately I did not loose a tweet.

HB27

george daniel wrote: »

ouch



I had a 555 that did the same thing,, fortunately I did not loose a tweet.

Yep. That's the varmint

heiney9

george daniel wrote: »

ouch



I had a 555 that did the same thing,, fortunately I did not loose a tweet.

Completely normal and not harmful for early Adcoms. Since they have no current limiting circuitry or inrush circuitry they pop slightly at power up. They buurrrrrrpppp or faaaaarrrrrtttttt at turn off if you turn the pre-amp off first.

My Adcom has been doing this for 22 years an I've had 3 sets of speakers in that time and had ZERO issue.

The reason they did this was to keep those nasty (I believe all that auto mute, current limiting, etc is completely unnecessary in a well built amp) signal altering components out of the signal path.

H9

HB27

heiney9 wrote: »

Completely normal and not harmful for early Adcoms. Since they have no current limiting circuitry or inrush circuitry they pop slightly at power up. They buurrrrrrpppp or faaaaarrrrrtttttt at turn off if you turn the pre-amp off first.

My Adcom has been doing this for 22 years an I've had 3 sets of speakers in that time and had ZERO issue.

The reason they did this was to keep those nasty (I believe all that auto mute, current limiting, etc is completely unnecessary in a well built amp) signal altering components out of the signal path.

H9

I figured out too late I'd better follow the start up and shut down sequences like a tube system with them AFTER I'd screwed up. Worst part of it was I knew better and flat out ignored my own better judgement. I'm a Fu..... KIA (know it all) at times.
Harry

ben62670

HB27 wrote: »

You can trip or damage a polyswitch just by turning the amp on or off if you have an amp that "pops or makes noise" on start up or shut down. I had an Adcom that did that on a regular basis and lost a tweeter even when the polyswitch tripped.
Harry

You may have a DC offset problem. Measure the DC offset by using a multimeter at the speaker terminals. With your amp you should have less than 24mv DC. Measure from the positive to the negative terminals with the source disconnected, and the speaker cables disconnected.
Ben

tugboat

ben62670 wrote: »

You may have a DC offset problem. Measure the DC offset by using a multimeter at the speaker terminals. With your amp you should have less than 24mv DC. Measure from the positive to the negative terminals with the source disconnected, and the speaker cables disconnected.
Ben

Don't forget to let the amp warm up for at least 10 minutes (I give it 30 minutes). This will give the bias voltage time to settle as it warms to operating temp.

ben62670

tugboat wrote: »

Don't forget to let the amp warm up for at least 10 minutes (I give it 30 minutes). This will give the bias voltage time to settle as it warms to operating temp.

Thanks I forgot that:o
Actually take a measurement taken right off the bat, and one after 10 minutes, and the last after 30. I would like to know if there is a burst at startup;)
Ben