Rotel RB-1090 Strange behavior with XLR/RCA connections

[MarkD]

Hello everyone!
I have old school Rotel RB-1090 amp connected to Eversolo A8 streaming pre-amp/dac. From day one I was using XLR balanced connection and yesterday decided to try RCA unbalanced, just to see if there are any difference that I could hear. So, I connected pair of RCA cables and a couple surprises came out. First, there is no volume level difference between XLR and RCA connection. It is absolutely identical. I was toggling switch at the back of Rotel 1090 between Balanced/Unbalanced circuit and no change in volume levels at all. Second and more strange, when I toggle the switch on Rotel 1090 to RCA (unbalanced) and left output from my Eversolo A8 on XLR balanced circuit, I can still hear the music playing, although at quite low level. Is that normal???
Or do I have issue in Rotel 1090 balanced/unbalanced circuits and sort of power leakage?

 

[Speedskater]

I think that it is functioning just as an old school design might.

 

[MarkD]

I think that it is functioning just as an old school design might.

Might be the case, not sure. But indeed strange behavior.

 

[AnalogSteph]

The balanced input appears to sport a gain of approx. -6.8 dB, which just about makes up for a 6 dB higher output level.
(RB-1090-2,3 schematic)

Looks like there's also a case of Pin 1 Problem, not sure how easily that could be resolved though.

 

[MarkD]

The balanced input appears to sport a gain of approx. -6.8 dB, which just about makes up for a 6 dB higher output level.
(RB-1090-2,3 schematic)
View attachment 436960
Looks like there's also a case of Pin 1 Problem, not sure how easily that could be resolved though.

Thanks for sharing this schematics! I don't hear any volume level differences between balanced and unbalanced connections. And this is strange, as my Eversolo A8 output 4.2V with balance and 2.1V with unbalanced connection.

 

[AnalogSteph]

Let me try this again:
The balanced output on the Eversolo is 6 dB louder than the unbalanced one.
The balanced input on the Rotel has an input gain about 6.6 dB lower than the unbalanced one.
Net, balanced should be about 0.6 dB quieter than unbalanced, which is barely detectable.

Got it now?

Basically, the Rotel was designed to make up for the gain of typical balanced output stages. If supplied with the same exact level, the balanced input would be a fair bit quieter than the unbalanced one.

 

[MarkD]

Let me try this again:
The balanced output on the Eversolo is 6 dB louder than the unbalanced one.
The balanced input on the Rotel has an input gain about 6.6 dB lower than the unbalanced one.
Net, balanced should be about 0.6 dB quieter than unbalanced, which is barely detectable.

Got it now?

Basically, the Rotel was designed to make up for the gain of typical balanced output stages. If supplied with the same exact level, the balanced input would be a fair bit quieter than the unbalanced one.

Thanks for clarifications, got it! I don't see the reason for using balanced connection now, as it seems Rotel 1090 is not fully balanced.

Do you know what could be the reason of sound leakage from balanced to unbalanced circuit? Do I need to service this unit because of that?

 

[Speedskater]

Thanks for clarifications, got it! I don't see the reason for using balanced connection now, as it seems Rotel 1090 is not fully balanced.

What does that mean?
There a three types of fully balanced outputs:
1] Transformer
2] Passive/Impedance
3] Active/Powered

 

[audio_tony]

There a three types of fully balanced outputs:
1] Transformer
2] Passive/Impedance
3] Active/Powered

The NAD1000 / 106 preamp has 'balanced' outputs but they're not balanced at all. So you could add this at 4 to your list...

 

[MarkD]

What does that mean?
There a three types of fully balanced outputs:
1] Transformer
2] Passive/Impedance
3] Active/Powered

"Rotel refers to "Balanced Design" in their literature, and the RB-1090 has balanced inputs, so one might think that the amp is a balanced design in the same way that products from companies like Balanced Audio Technology are balanced. It's not. Like virtually all moderately priced amplifiers with balanced input connectors, the Rotel is what some call "pseudo-balanced," in that the circuitry itself is unbalanced; balanced input signals are converted to unbalanced internally"

The full read and measurtements here - https://www.stereophile.com/solidpoweramps/100rotel/index.html

In such cases where balanced signal is internally converted to unbalanced, it is better to use unbalabce connection between preamp and amp, just to avoid a number of convertions.

 

[AnalogSteph]

I don't see the reason for using balanced connection now, as it seems Rotel 1090 is not fully balanced.

Huh? No idea how you came to this kind of conclusion.

The input should still provide some CMRR, even if this may be limited to less than what the input stage could do due to the Pin 1 Problem.

Do you know what could be the reason of sound leakage from balanced to unbalanced circuit?

Good question. Maybe a bit of capacitive leakage (and/or switch cleaner residue) in the bal/unbal switch. When you've got a 34 kOhm input impedance following, it doesn't take that much. (There might be a pinch of shared ground return impedance in the mix as well, although with source impedances in the tens of kOhms this would have to be pretty severe for any crosstalk to speak of to occur.)

EDIT:

in that the circuitry itself is unbalanced; balanced input signals are converted to unbalanced internally"

Well, that is true.

In such cases where balanced signal is internally converted to unbalanced, it is better to use unbalabce connection between preamp and amp, just to avoid a number of convertions.

That's completely wrong. It fundamentally is perfectly acceptable to use unbalanced circuitry internally and only interface with balanced connections to the outside world. (Performance can still be first-rate, see e.g. the Topping L50 headphone amp.) In the case of an IEC Class I device like this amp, this can actually be a very good idea when it comes to avoiding ground loops.

 

[MarkD]

Huh? No idea how you came to this kind of conclusion.

From the link above and couple of other reviewes of this amp.

 

[MarkD]

to the Pin 1 Problem

I didn't get what is the problem with Pin 1. Can you please elaborate on this a little bit?

Huh? No idea how you came to this kind of conclusion.

The input should still provide some CMRR, even if this may be limited to less than what the input stage could do due to the Pin 1 Problem.


Good question. Maybe a bit of capacitive leakage (and/or switch cleaner residue) in the bal/unbal switch. When you've got a 34 kOhm input impedance following, it doesn't take that much. (There might be a pinch of shared ground return impedance in the mix as well, although with source impedances in the tens of kOhms this would have to be pretty severe for any crosstalk to speak of to occur.)

So, its time for service then. Well, at least inside clean up wouldn't hurt.

 

[somebodyelse]

The NAD1000 / 106 preamp has 'balanced' outputs but they're not balanced at all. So you could add this at 4 to your list...

View attachment 437410

R343/344 look like an approximation to Z-out, so an attempt at impedance balanced. It may not be a great match, but it's better than nothing.

 

[somebodyelse]

I didn't get what is the problem with Pin 1. Can you please elaborate on this a little bit?

It doesn't follow AES48 as Pin 1 is not connected to chassis at the connector. As a result there's a good chance that ground-related noise will be coupled into the signal. In other words it may hum if there's a ground loop, which wouldn't happen if it was implemented properly. This is generally known as a 'Pin 1 Problem'.

 

[MarkD]

It doesn't follow AES48 as Pin 1 is not connected to chassis at the connector. As a result there's a good chance that ground-related noise will be coupled into the signal. In other words it may hum if there's a ground loop, which wouldn't happen if it was implemented properly. This is generally known as a 'Pin 1 Problem'.

Thanks for explanation. Fortunately, there is no any hum coming out from the amp. My only concern now is that it looks like my Eversolo A8 connected directly to Rotel 1090 amp cannot give full power to the amp, as I can easily raise volume to -8db, while 0db is a maximum. It like I am close to the maximum volume of 1090, which is 380W into 8 and 700 into 4omhs. Speakers are quite sensitive too, at 93db (Focal Electra 1038 BE2).

 

[AnalogSteph]

It doesn't follow AES48 as Pin 1 is not connected to chassis at the connector. As a result there's a good chance that ground-related noise will be coupled into the signal. In other words it may hum if there's a ground loop, which wouldn't happen if it was implemented properly. This is generally known as a 'Pin 1 Problem'.

Exactly. This is a rather common constructive problem pre-2005 (which saw the release of AES48-2005), and even more so pre-1995. It has yet to be fully eradicated in some kinds of devices like inexpensive active monitors... doing it properly may complicate assembly.

EDIT: Oops, ninja'd.

My only concern now is that it looks like my Eversolo A8 connected directly to Rotel 1090 amp cannot give full power to the amp, as I can easily raise volume to -8db, while 0db is a maximum. [...] Speakers are quite sensitive too, at 93db (Focal Electra 1038 BE2).

That's odd. You may need to investigate why you seem to have such low digital levels. At 28.8 dB unbal / 22 dB bal, the amplifier sports a decent amount of gain; things should go pretty loud like that.

 

[MarkD]

In the case of an IEC Class I device like this amp, this can actually be a very good idea when it comes to avoiding ground loops.

Thanks. I am not good at amp's schematics and what is the best way to go in such cases, so I really appreciate your inputs and recommendations!

 

[MarkD]

That's odd. You may need to investigate why you seem to have such low digital levels. At 28.8 dB unbal / 22 dB bal, the amplifier sports a decent amount of gain; things should go pretty loud like tha

My case is not isolated, as I found a number of complaints on Zidoo (Eversolo) forum about the same issue. Unfortunately, no any response from Eversolo. Today I tried my previous preamp which is Rotel RC1572 MK2 and I was blown away by the difference both in volume levels and quality! I am not an audiophile, but even I was stroke by such a huge difference even from the 1st second as I started to play music. Something is definitely wrong in pre-amp section of Eversolo A8.

 

[Speedskater]

The NAD1000 / 106 preamp has 'balanced' outputs but they're not balanced at all. So you could add this at 4 to your list...

While those are XLR output connectors, they are not balanced outputs.