XLR to RCA to XLR


I decided to try the KUBE (KEF's equalizer) with KEF R104.2s in a balanced system. By careful routing of cables my system is silent to an ear at the speaker. But the KUBE has only RCA connectors, and adding it to my system introduced a sizzle, a kind of HF hum. I removed the KUBE, and silence returned. Is there any way of going from XLR to RCA that might avoid the noise?

Balanced lines (about 15') run from a Cary 11a processor to a Proceed Amp 3 that sits near the R104.2s. The third line to the center speaker did not run through the KUBE and thus remained balanced.

db
Ag insider logo xs@2xdbphd
The KUBE 200 has LF and HF contour adjustments, plus bypass and tape monitor loop switches. The HF contour does provide a subtle boost or cut, but I haven't messed with it much. The bypass definitely works, although I don't know how "true" of a bypass it is. Here is a link to the manual if you're interested:
http://www.kef.com/uploads/files/en/museum_pdf/80s/Reference_Series_Model_KUBE_200_Installation_Manual.pdf
Right now, I am using the KUBE at low to moderate volume and bypassing it if I go higher.
Unlike the KUBEs that are integral to the design of the speaker, I think the only purpose of the 200 KUBE is bass boost. I had a 200 KUBE at one time and I don't recall if it had a button you can push to remove the boost circuit, but I suspect it may not. The 102 and 107/2 KUBEs I use both have such a button. I use those KUBEs without the added boost.

db
Since my last post, I was lucky enough to speak with Bill Whitlock of Jensen, who is both very knowledgeable and willing to give thoughts and suggestions about my situation. He observed that the gain mismatch I am getting would be best remedied by somehow getting more output from the KUBE. I sent him schematics of the KUBE and he replied back very quickly, instructing to change out 2 resistors in the variable output circuit for lower values, which would extend the gain by a few dB. He said that I could stick with the RCA to XLR cables the way I had them wired, but suggested that if I got any noise, I should try combining the shield with the black conductor at the RCA end. If no luck there, then I should look into getting a PI-2RX, etc.

I have yet to do this because I haven't gotten the parts, but in the meantime, I figured I'd had a go with db's suggestion above and just try it for myself. I hooked it up, non-biamped, and gave it a go.

I have to say I really don't notice any discernible difference in the HF with the KUBE inline. It may be that my gear is just not good enough to expose anything, but I went back and forth a few times and I'll be damned if I heard any detrimental effects on the mids or highs.

The first difference I did notice is, I have less gain from using the RCA to XLR outputs from the KUBE to my amp. I have to crank the preamp up a bit more, but I thankfully don't have any hiss, buzz or hum. I'm going to live with this setup until I get the parts I need to mod the KUBE for biamping, or I may just make another set of RCA to XLR cables and forget the balanced preamp output altogether.

Now, as far as using the KUBE...and forgive me if I am taking this thread off into the weeds... The whole point of getting one (for me) was to see just how well it could extend the low end. As db pointed out, the KUBE is not "required" for the 105/3, but it's an option if you feel you're missing out on the lower octaves. I can say that it does indeed work as advertised, but I think it may be at the expense of overall SPL and some very obvious strain on the woofers.

A few observations:

- With the LF and HF controls set at the center notch (which I would think would be flat), there was a noticeable bass boost when bypassing and un-bypassing the KUBE, so it seems those knob positions are not flat, or the same as bypass.

- When engaging the KUBE at a decent volume with the LF contour slightly increased and some moderately bass-heavy material, the bottom woofer's movement started to scare me...

- The KUBE seems to not only extend the bass, but also change the tone of the bass, if that makes any sense. I noticed this regardless of whether the LF knob was centered or not. It almost seems like I lose a tiny bit of upper mid-bass as soon as the KUBE is un-bypassed.

Overall, I think the KUBE on the 105/3 is a nice option if you want to get a little more oomph, but only at low to moderate volumes. For "critical" listening or filling the house with music, I would be wary of using it for fear of damage. IMO, the KUBE tries to make the 105/3 woofers do something for which they were not designed.
Rdeckard wrote, "I only want the KUBE on the LF in a biamp'd config. I have read that even if you do not adjust the HF contour on the KUBE, it can still muddy up the wonderful midrange on these speakers just by being in the signal path, which is probably another debate altogether..."

According to the KEF museum, the KUBE is just an option for the 105/3, not integral to the design as it is for the 107/2. Why not just try it between the preamp and amp, and judge for yourself whether it muddies the midrange?

The 107/2 KUBE doesn't muddy the sound of my spectacular 107/2s that are bi-amped between the amps and the speakers.

db
Rdeckard, I took a look at the manual for the Kube 200, and I see that its level control provides a maximum boost of 5 db for its variable output, relative to the level of the fixed output. Your findings seem consistent with that.

The approach you are considering would work, except that it may be that a PI-2RX should be used between the Kube and the power amp, rather a PI-2XR connected "backwards".

HOWEVER, in the light of Grant's (Tvad's) comment above that ...
Jensen makes 1:1, 1:2, 1:4, 2:1, 4:1 transformers, and I suspect any of these could be built into their XLR>RCA or RCA>XLR converter boxes at a customer's request.
... a call to Jensen would seem to be in order, to explore additional possibilities.

Those possibilities would include stepping up the signal level by 1:2 (6 db) between the Kube and the power amp; or stepping down the mid/hi frequencies by inserting into the XLR connection between preamp and power amp a PI-2XX (XLR ins and outs) specially configured with a 2:1 stepdown transformer (with no transformers being used in the low frequency part of the signal path, unless one is needed to eliminate groundloop hum).

Looking at all of the numbers that are involved I don't see anything that would necessarily rule out any of those alternatives. But my instinct, assuming that you are not presently using your volume control near the upper end of its range, would be that it is generally better to step down than to step up, given the cautions I expressed yesterday in my post about the Rane unit. And also given the very high degree of transparency Tvad and others have reported the Jensen transformers as providing.

Hope that helps. Regards,
-- Al
Thanks everyone for all the helpful info!

Here is more info on my setup and my latest experiment...

My preamp has both RCA and XLR outputs, which can be used simultaneously.

I only want the KUBE on the LF in a biamp'd config. I have read that even if you do not adjust the HF contour on the KUBE, it can still muddy up the wonderful midrange on these speakers just by being in the signal path, which is probably another debate altogether...

Anyway, I am feeding the XLR outputs of the preamp straight into the the XLR inputs of my amp for the HF.

I then figured I'd use the RCA outputs of the preamp to feed the KUBE, and then the KUBE output would go to a second set of amp inputs (it's a 7 channel amp).

Last night, I tried this with just RCA to XLR cables on the output of the KUBE to my amp using cable #17 on this page: http://www.rane.com/note110.html

It "worked", but the output from the LF side was significantly lower, even when using the variable output of the KUBE and cranking the pot on the back.

I sent a 60Hz tone through the system at low volume and with a volt meter, measured .40v on the HF speaker terminals and .30v on the LF speaker terminals. I'm sure this is not very scientific, but I wanted to somehow confirm what my ears were hearing.

I guess I have voltage mismatches by using the unbalanced outputs from my pre to the KUBE compared to the straight balanced connections for the HF.

Now I am thinking I should NOT use the unbalanced preamp output, but rather split the balanced preamp outputs with XLR Y-adapters (this is sounding less and less high-end as I write), feed one directly to the amp for the HF, then send the other to a Jensen PI-2XR to feed the KUBE.

The unbalanced output of the KUBE would then go through another PI-2XR to the second pair of balanced amp inputs.

I figure in this setup I should (hopefully) have similar gain/voltage at both pairs of amp inputs when the KUBE is inline (and bypassed).

Here are some quick specs on the gear:

KUBE 200:

Input Impedance 51 k ohms.
Maximum Input Voltage >1.9 V rms above 20 Hz.
>8 V rms above 50 Hz.
Maximum Output Voltage >5 V rms.
Output Impedance 100 ohms.
Output Noise Level <6 uV A-weighted.
Signal-to-Noise Ratio >104 dBA ref 1 V rms.
Distortion <0.005 % for 1 V rms input @ 1 kHz.

Preamp:

Unbalanced line-level outputs:
Maximum Output Level: 4 Vrms
Output Impedance: 100 ohms

Balanced line-level outputs:
Maximum Output Level: 8 Vrms
Output Impedance: 200 ohms

Amp:

Balanced line-level audio inputs:
Input Impedance: 24k Ohms balanced
Maximum Input Level: 4 Vrms balanced

Also, since it just says balanced and unbalanced "line-level" I don't think I need to worry about +4 to -10 conversion, etc.

Apologies for the long post. Any advice or suggestions at this point (or a slap in the face if I'm doing something stupid) would be most appreciated!
01-23-13: Rdeckard
What's confusing is, it seems balanced doesn't always mean "pro" or +4 and unbalanced doesn't always mean "consumer" or -10. What does most high-end balanced home audio (not pro) gear operate at, +4 or -10?
The reference to unbalanced signal levels on consumer equipment being around -10 dbV lost whatever meaningfulness it may have once had when the CD medium was introduced, which was standardized with a full scale (maximum) level of 2 volts, for unbalanced outputs. 2 volts is +6 dbV, or +8.2 dbu. Some digitally based equipment will exceed even those numbers significantly, while phono stage outputs will usually be considerably less.

Each situation needs to be analyzed individually, although it is more common these days to have too much signal and too much gain, rather than too little, regardless of whether the connections are balanced or unbalanced.

Grant (Tvad), thanks for your always knowledgeable inputs.

Regards,
-- Al
Post removed 
The Rane piece looks good. I like that it's totally passive. I wonder if it would work well just for balanced to unbalanced when -10 to +4 conversion is not a concern.
What's confusing is, it seems balanced doesn't always mean "pro" or +4 and unbalanced doesn't always mean "consumer" or -10. What does most high-end balanced home audio (not pro) gear operate at, +4 or -10?
Hi Jedinite,

Regarding the Rane BB44X, which I see sells at B&H for $279, based on a quick Google search I didn't find much that would shed light on the sound quality it would provide in a high-end home audio system.

Looking at the datasheet, though, a concern would be that while a statement is made that the unit provides "wide bandwidth," the only spec that is provided is that frequency response is 20 Hz to 20 kHz +/- 2 db. The 2 db tolerance is very uncomfortable, IMO, and the fact that no margin is indicated relative to the audible frequency range is as well. Without a margin of 5 times or more (i.e. 4 Hz to 100 kHz), and preferably even approaching a factor of 10, I would not feel comfortable that audibly significant phase shifts wouldn't occur within the 20 Hz to 20 kHz range.

In contrast, Steve McCormack's unit is specified as having a 3 db bandwidth of 0.5 Hz to 150 kHz, and the frequency response graph shown in the datasheet for the Jensen PC2XR shows it as being in a similar ballpark.

Also, some cautions that would apply if any of these level converters were used to step UP "consumer level" to "pro level":

First, using one of these devices in that direction will increase the impedance seen "looking back" from the destination device toward the signal source by a factor of about 16 (corresponding to the square of the nominally 4:1 turns ratio), and will decrease the load impedance seen by the signal source by that same factor of 16. That will result in impedance incompatibilities in many cases.

Second, at 20 Hz the unbalanced input of the Rane device will reach its overload point at only 1.68 volts, with the Jensen PC2XR being only slightly better in that respect. So, as might be expected, one would not want to use one of these transformers to step UP the outputs of unbalanced signal sources which provide the 2 volt kinds of levels that are typical these days, especially with digital sources. And chances are that other components in the signal path would overload as well, if that were done.

These two cautions do NOT apply to 1:1 transformers, such as the Jensen PI models, and do NOT apply to the 4:1 level converter devices when used as step-DOWN transformers, i.e., going FROM "pro levels" TO "consumer levels."

Regards,
-- Al
Thanks Tvad and Almarg for the info. I've since looked into it a bit more with regards to the input and output voltages I'm dealing with and I may be OK just using XLR to RCA cables with the shields floating at the RCA ends. If that doesn't yield the results I want, I will pick up some transformers.
I now have Jensen transformers for the 107/2 and 102 KUBEs, plus a spare. I found that sound quality was better with XLR outputs from the transformers to the amps. Since my May '12 post, I've replaced the Proceed Amp 2 & 3 with Proceed HPA 2 & 3 (500 watts peer channel into 4 ohms).
Hi Al

Thanks for posting about SMc Audio's Flex-Connect. I believe they called it the Interocitor before. For mixing consumer and pro audio what about the Rane Balance Buddy 44x?

http://www.rane.com/pdf/bb44xdat.pdf

Not the prettiest looking piece of gear but could be available used at a decent price.
01-21-13: Rdeckard
Will these work if mixing consumer audio (-10dBV) and pro audio (+4dBu) gear? In other words, do they provide any level-matching?
The PI models will function in that situation, but they will only provide a very slight amount of attenuation (corresponding to their "insertion loss," which under the test conditions shown in the datasheet is 1.6 db typical, 2.0 db max).

Assuming that what you want to do is to reduce the signal level, what you would probably want is their model PC2-XR, which provides about 14 db of attenuation. A phone call to them would be a good idea, though, to confirm its suitability to your specific application.

Also, at a considerably higher price point Steve McCormack of SMc Audio offers a transformer-based product called Flex-Connect, which provides 12 db of attenuation, together with various other features. I have no specific knowledge of it, but I know that his VRE-1 preamplifier, which utilizes coupling transformers, has an exceptionally good reputation.

Regards,
-- Al
Post removed 
Gotcha. Will these work if mixing consumer audio (-10dBV) and pro audio (+4dBu) gear? In other words, do they provide any level-matching? Thanks.
Rdeckard, see the part number table at the bottom of this page. Information shown in the PI-2XX datasheet is applicable to the versions that provide either RCA or XLR inputs and/or outputs, as well as either 1 or 2 channels. The various configurations are differentiated in accordance with the part number definitions shown in the table.

Regards,
-- Al
I am in the same boat with a KUBE 200 for a pair of KEF 105/3's. I need to go from balanced preamp outputs to the unbalanced KUBE intputs, then the unbalanced KUBE outputs to balanced amp inputs.
I looked at the Jensen site, but it seems the PI-XR is not listed, unless you have to order one in a certain configuration. Any other alternatives out there as well? Thanks.
Followup: I've installed a 107/2 KUBE using ISO MAX transformers. It provides LF extension to 20 Hz, although the 107/2s are not LF shy, and it seems to provide smoothing in the mid crossover regions. I changed the SACD output from the Oppo 95 to DSD rather than PCM, because the mains do not need to be supplemented by the subs.

Now I need to buy an ISO MAX for the 102 KUBE for the center channel.

db
Sorry, Almarg, I don't, and I don't have a multimeter.

Tvad somewhat reluctantly sold his pair of ISOMAX transformers to me. Now I have to find a KUBE for the KEF 107.2s I'm buying.

db
Kudos to Tvad!!

DB, just out of curiosity do you know if the adapter you were using at the INPUT to the KUBE (the XLR-female to RCA-male adapter) is one of those that shorts XLR pin 3 to XLR pin 1?

If you happen to have a multimeter that can be easily determined, of course.

Regards,
-- Al
Tvad's ISOMAX transformers arrived today. I tried them using the KEF Model 200 KUBE that had previously caused the sizzle. The result was flawless: Silence at the speaker, and, even though designed for use with a different speaker, the KUBE provided substantial LF extension. I played the Ondine SACD of Saint Seans 3rd symphony. Pedal notes that were inaudible with the unaided 104.2s became audible.

Had I known the KUBE was so effective, I would have looked for one tailored for the 104.2s. Anyone have a lead on a KUBE for KEF 107.2s?

db
Mine was the winning bid on a Model 102 KUBE to use with vintage KEF 102 speakers I use for center channel. I've arranged with Tvad to try his Jenson ISOMAX transformers in the loop. I'll report back.

db
Jedinite, thanks very much. Regarding ferrites, they will not help a hum problem, as they are only effective at radio frequencies (and higher).
05-29-12: Dbphd
I used a female XLR to male RCA adapter into the KUBE from the XLR line out of a Cary 11a and a male RCA to male XLR adapter to the line into a Proceed Amp 3.
Most adapters of either type short the signal on XLR pin 3 to ground (pin 1). That is what should happen when adapting to an XLR input. However, IMO it makes absolutely no sense for XLR-female adapters, that would be used on outputs, to be designed that way. As I understand it, Cardas adapters are one of the few types that (correctly) do not short pin 3 to pin 1 on XLR female adapters.

This is just a speculative guess, for which I can't formulate a precise explanation, but if you are not using Cardas adapters I'm wondering if the noise may be somehow resulting from an inability by the Cary 11a to handle having its output on pin 3 shorted to ground. Many components can handle that ok, but some cannot, this being one example. I note that you referred to "a sizzle, a kind of HF hum," while the poster in that thread referred to "buzzing," those being descriptions that are not all that dissimilar. And I suspect that the fact that the short to ground is at the far end of the cable, at the input to the KUBE, may not be helping matters.

Does the noise still occur if the Cary is turned off or disconnected? Is it sensitive to the volume control setting?

Regards,
-- Al
Jedinite,

As I understand it, and my understanding may be incorrect, the use of a KUBE with the 104.2s is optional -- it only extends LF. In some applications, e.g. 107.2, the KUBE also tailors audio above the usual sub crossover. Thus, if you use a sub with your 104.2s, you may not need a KUBE.

The KEF museum is a treasure trove of information where you can download the installation instructions for the 104.2 as well as a nifty brochure that shows a cutaway of the speaker.

I have KEF 102s, 102.2s, 104.2s, and plan to buy 107.2s, so I've done lots of downloading and reading.

db
Hi Dbphd

Thanks for posting this thread. I may be facing a similar issue soon. I'm have a pair of KEF 104/2s and was able to score the 104/2 Kube for them. Nothing is hooked up at the moment due to renovations being done in my house. My amp and preamp combo have XLR inputs and outputs. I was going to put the Kube in the signal path between preamp and amp and use XLR adapters for at the preamp and amp sides. I hooked up a temporary system with the KEF Kube in place and heard a faint hum from my speakers. It was only when I was close to them though. I don't think the issue is you having a KEF Kube 102/2 with your 104/2s. It may just be the KEF Kube itself. I routed the AC Adapter and it cords as far away as I could from the unbalanced cables but there was still a hum.

Al. Many thanks for your contribution as well. Always very sage advice.

Would ferrite choke cores on the plug and umbilical cord ends help reduce the hum? I was thinking about putting choke cores near where the wires meet AC adapters as well where the umbilical cord plugs in along with where the 2 prong plug is.

Here are some pics of the KEF Kube. It looks the same for the 104/2, 102/2 and 103/3 speakers.

KEF Kube rear
http://img.2dehands.be/f/normal/120814494_3-kef-104-2-kube-inbegrepen.jpg

KEF Kube front
http://img.2dehands.be/f/normal/120814494_2-kef-104-2-kube-inbegrepen.jpg

KEF Kube AC Adapter
http://p2.la-img.com/236/423/197124_1_l.jpg

If the hum still persists I may just sell the KEF Kube.
Al,

The KUBE has a 2-prong AC plug.

I used a female XLR to male RCA adapter into the KUBE from the XLR line out of a Cary 11a and a male RCA to male XLR adapter to the line into a Proceed Amp 3.

Turns out I was using a KUBE designed for the KEF 102.2s I have in another system that has short lines between a Proceed PAV.PDSD and Amp 2, so unbalanced is no problem. But I'm negotiating the purchase of a pair of KEF 107.2s to replace the 104.2s, and a 107.2 KUBE is recommended by KEF.
Assuming that the KUBE is connected between the processor and the amp, and if it has a 3-prong AC power plug, chances are that the noise is being caused by a ground loop between the KUBE and one or both of the other two components.

To verify that, temporarily isolate the AC safety ground pin on the KUBE's power plug by connecting it via a 3-prong to 2-prong adapter (a "cheater" plug), with the adapter's ground lug or ground wire left unconnected.

That will break any ground loops between the KUBE and the components it is connected to. If that eliminates the noise, remove the adapter from the KUBE's plug and use it on the plugs of each of the other two components, one at a time. That will enable you to determine which interconnection is causing the problematic ground loop (although the answer might be both).

Post back your results, and also see the following references:

http://www.jensen-transformers.com/an/an003.pdf.

http://www.jensen-transformers.com/an/an004.pdf.

Also, if you are connecting the KUBE to the power amp by means of an unbalanced RCA cable followed by an RCA-female to XLR-male adapter at the amp's input, you may be able to improve the noise by instead using an RCA-male to XLR-male adapter at the KUBE's output followed by a balanced XLR cable. That would be electrically equivalent to the approach shown in Figure 2.1 of the first reference. Although as indicated in the reference that approach is less ideal than a well implemented transformer-based approach.

Regards,
-- Al
Post removed 
Hi Dbphd,

FWIW, I had the same problem running from my SE preamp to XLR only amp. I ended up trying a Sablon Audio interconnect (specially terminated RCA-XLR), and i went back to silent. the designer said it could have been luck. (I have Transparent Audio Ref in the rest of my system). sablonaudio.com

BTW, excellent cables in the upper TA league. 30-day trial period.