WolfX700 Measurements of Topping Pre90/Ext90 Preamplifier/ Extender

[MusicNBeer]

The performance of these passive types is only good near max or near nil. You can read Benchmark's description for their LA4 so you don't need to take my word for it. Passive types always have major trade offs between input impedance, output impedance, noise performance, interference and distortion.

I disagree with your assessment.

The input impdeance is near constant at whatever the pot value is until you get to max where the input impedance reduces to pot in parallel with load. Output impedance is max at near -6dB relative to max where it's pot value divided by 4. For most of the attenuation range, output impedance is way lower than this.

I analyzed all this before settling on 10K pot. Even 2.5K Ohm output impedance is no problem with short cable run. It works great with essentially no added distortion.

 

[JohnYang1997]

I disagree with your assessment.

The input impdeance is near constant at whatever the pot value is until you get to max where the input impedance reduces to pot in parallel with load. Output impedance is max at near -6dB relative to max where it's pot value divided by 4. For most of the attenuation range, output impedance is way lower than this.

I analyzed all this before settling on 10K pot. Even 2.5K Ohm output impedance is no problem with short cable run. It works great with essentially no added distortion.

Sure it's fine with short cable.

 

[Cortes]

The performance of these passive types is only good near max or near nil. You can read Benchmark's description for their LA4 so you don't need to take my word for it. Passive types always have major trade offs between input impedance, output impedance, noise performance, interference and distortion.

But where is the noise, interference and distorsions in the measurements I linked above?. All occurs at the resolution limit of the AP. Sorry, I'm just trying to figure out.

 

[MusicNBeer]

But where is the noise, interference and distorsions in the measurements I linked above?. All occurs at the resolution limit of the AP. Sorry, I'm just trying to figure out.

There is no harmonic distortion, it's all resistors. There is noise but it's probably 50 dB below the noise floor of most rooms. Interference? I don't understand that one. Interference is a cabling problem, not metal box problem.

I do freely admit, the usefulness of these super high precision passive preamps has been reduced hugely by the superb DACs now with digital volume control will vanishing noise floor and distortion.

 

[okok]

snr? under * vrms

 

[YSC]

I disagree with your assessment.

The input impdeance is near constant at whatever the pot value is until you get to max where the input impedance reduces to pot in parallel with load. Output impedance is max at near -6dB relative to max where it's pot value divided by 4. For most of the attenuation range, output impedance is way lower than this.

I analyzed all this before settling on 10K pot. Even 2.5K Ohm output impedance is no problem with short cable run. It works great with essentially no added distortion.

Sorry to interrupt, but does this apply to the in series (serial) type stepped attenuator? I actually wonders what will be the output impedance is going from mid attenuation downwards, as if we use new max volume level there isn't a reason we would want an expensive passive preamp?

 

[JohnYang1997]

Resistor ladder usually has low distortion because the voltage is distributed across all those resistors. However there are types of stepped attenuation topologies which supposedly reduce resistors in the signal path for more "transparency" (which is BS). The shunt, current control or whatever the types those are will expose all the input voltage across 2 or 4 resistors which can cause distortion due to VCR (voltage coefficient). Some resistors are better some are worse. The differences of distortion can be up to 80dB if we include carbon composite type, and 30-40dB for other types. So yeah resistors have distortion.

 

[JohnYang1997]

Sorry to interrupt, but does this apply to the in series (serial) type stepped attenuator? I actually wonders what will be the output impedance is going from mid attenuation downwards, as if we use new max volume level there isn't a reason we would want an expensive passive preamp?

There are different topologies. Common one is like regular potentiometer. Constant input resistance, varied output resistance.
The output impedance will be the two sections of the resistance in parallel. 50% position = 5k//5k = 2.5k. For lower position the output impedance will become lower which is good.

 

[YSC]

There are different topologies. Common one is like regular potentiometer. Constant input resistance, varied output resistance.
The output impedance will be the two sections of the resistance in parallel. 50% position = 5k//5k = 2.5k. For lower position the output impedance will become lower which is good.

ic, so both the Khozmo and goldpoint standard serial topology passives have max resistance in mid point (50% volume) and goes lower either way? and even for 2.5k ohm out and with a 10k in for quite some chip power amp have don't seems to have a significant effect?

 

[JohnYang1997]

But where is the noise, interference and distorsions in the measurements I linked above?. All occurs at the resolution limit of the AP. Sorry, I'm just trying to figure out.

The writer wanted to bust the myths hence the result fits what he expect.
What his measurements shown are correct. However here's where the differences are: For high value attenuators output impedance will be high in certain range. He's testing under -20db which is way lower than the range. But even at that range you will see difference in noise if you add 20- 40dB amplifier afterwards, which is normally what you have with a power amp. No I'm not saying you have to buy the best preamp to enjoy music. Far from it. The point is that you either get better preamps or you can stick with digital volume for essentially perfect real world experience. Also balanced connection has better rejection to noise to begin with. Maybe I was wrong categorize the preamp you linked(which appears to be balanced only) to other passive preamps. But you get the idea.

 

[JohnYang1997]

ic, so both the Khozmo and goldpoint standard serial topology passives have max resistance in mid point (50% volume) and goes lower either way? and even for 2.5k ohm out and with a 10k in for quite some chip power amp have don't seems to have a significant effect?

I would say higher output impedance is suspectable for noise pick up and interference. Regular power amps have input referred noise of about 1k to 2k ohm resistor. I would say 2.5k(corresponding to 10k pot/attenuator) is as high I would go, 12.5k (50k pot/ attenuator) is certainly too much to me.

 

[Billy Budapest]

I'm back... Today I measured Pre90 Preamplifier and Ext90 Extender from Topping.

This is a set of equipment that I have been looking forward to for a long time...I have received the prototype a few months ago, but my audio analyzer was lent out. When the analyzer came back, unfortunately, the review time lock in the prototype had been activated-the entire device was locked and unusable.

I contacted Topping and got a "retail version" of this measurement.

Anyway, look at things first

View attachment 94302
View attachment 94303


Previously, Topping's A90 could be used as a preamplifier, but it was limited by the basic design (no remote control, only the volume can be adjusted manually). The preamp function of A90 is just better than nothing for me. (But it might be useful for desktop users) What I need is a preamplifier for "living room" or "audio room". So when I knew that Topping was going to develop a preamplifier with remote control, I looked forward to it and made some suggestions of my own.

First, it is almost completely transparent, almost as transparent as the signal cable measured by APx555.
Then, it must have enough volume adjustment range and very good channel balance:
View attachment 94304

In addition, it can expand more input ports (actually, I originally recommended daisy-chain connection so that you can connect multiple extenders at the same time... but Topping rejected this requirement). This is for users with multiple audio sources The user is more friendly.

Then it is very important to use experience! First of all, there must be a volume memory function (different signal sources have different output levels, and a volume memory function can ensure a good experience when switching between different input sources.)

But is the volume memory enough? No, that's not enough... Let's assume a scenario (actually I have actually encountered it), when a source is not turned on or there is no sound from the source, maybe I unknowingly turned up the volume. Then I found that the source was not turned on, so I switched to another source. Two days later, I turned on the source that was not turned on before, and then switched. Due to the volume memory function...maybe I was deaf right away...so at my request, Topping added a safe volume function:
View attachment 94306
View attachment 94307

The logic of this function is that when the memory volume is called, if it is higher than the set safety level, it will be muted first, let you press OK to call it or press VOL + to gradually adjust it.

Okay, stop talking nonsense... Measurement is the most important thing...


XLRin 4V Dashboard
View attachment 94311


XLRin 5V Dashboard
View attachment 94312

RCA in Dashboard
View attachment 94313

EXT XLR Dashboard
View attachment 94314


THD+N Vs Freq
View attachment 94315

FR:View attachment 94323

Multitone
View attachment 94318

And Output impedance:
View attachment 94335

I know that for many users, the necessity of preamplifiers is debatable. But I believe that there is a certain market for pre-amplifiers with multiple inputs, good functions and user experience, especially if it has excellent performance and inexpensive prices.

This is exactly what I want.

Can you repost the first picture without the L7 AudioLab logo obscuring the extender unit? Thanks.

 

[JohnYang1997]

Can you repost the first picture without the L7 AudioLab logo obscuring the extender unit? Thanks.

 

[YSC]

I would say higher output impedance is suspectable for noise pick up and interference. Regular power amps have input referred noise of about 1k to 2k ohm resistor. I would say 2.5k(corresponding to 10k pot/attenuator) is as high I would go, 12.5k (50k pot/ attenuator) is certainly too much to me.

So it seems the 10k version we likely won’t have any degradation to SINAD with most active speakers or power amps and is a nice option to use?

 

[tmtomh]

Can someone explain, for us idiots , what defines a line-level source component that is "robust" enough to mate well with the Pre90's 2k ohm input impedance? I have seen only "modern" equated with "robust" here, plus an example of an Accuphase tuner.

To put it another way, it seems to me that a lot of modern line-level source components do not include output impedance in their specifcations. For example, my Oppo UDP-205 lists output impedance only for the headphone output, not the line-level outputs. So how would I go about determining whether or not that component would work okay with the Pre90?

 

[JohnYang1997]

Can someone explain, for us idiots , what defines a line-level source component that is "robust" enough to mate well with the Pre90's 2k ohm input impedance? I have seen only "modern" equated with "robust" here, plus an example of an Accuphase tuner.

To put it another way, it seems to me that a lot of modern line-level source components do not include output impedance in their specifcations. For example, my Oppo UDP-205 lists output impedance only for the headphone output, not the line-level outputs. So how would I go about determining whether or not that component would work okay with the Pre90?

I think I have explained in one of the previous posts discussing all cases that may have issues. So UDP-205 is fine regardless of what output impedance of that thing actually is. You only need to turn the preamp up a bit in the worst case scenario.

 

[YSC]

I would say higher output impedance is suspectable for noise pick up and interference. Regular power amps have input referred noise of about 1k to 2k ohm resistor. I would say 2.5k(corresponding to 10k pot/attenuator) is as high I would go, 12.5k (50k pot/ attenuator) is certainly too much to me.

Or if I have a resistor used in series list can we have a formula to deduce the output impedance? Nice if I can learn about this, Googling makes me confused

 

[JohnYang1997]

Or if I have a resistor used in series list can we have a formula to deduce the output impedance? Nice if I can learn about this, Googling makes me confused

There really isn't. Best way is to add a buffer. But then it's something else. Or you can reduce the input impedance further. 2k is quite low on the look but usable.

 

[YSC]

There really isn't. Best way is to add a buffer. But then it's something else. Or you can reduce the input impedance further. 2k is quite low on the look but usable.

O, so even if having a resistor value we can't compute theoretical value of output impedance, I thought it was my stupidity resulting in this

 

[JohnYang1997]

O, so even if having a resistor value we can't compute theoretical value of output impedance, I thought it was my stupidity resulting in this

Output impedance changes across the range. 0 at maximum and nil position and 1/4 of the resistance at 50% position. And in between else where. You can calculate it but there's no way to minimize it without changing topologies or adding a buffer.