More Disruptive Schiit (Vidar)

[esldude]

Vidar looks like a pretty nice amp for the money etc.  OTOH, why is this disruptive?

 

For one comparison I don't see a big huge advantage to the very similarly spec'd, less expesnive, more powerful amplifier like this one here:

 

https://emotiva.com/products/amplifiers/300

 

Does the microprocessor in this one also cause odd behaviour like the Ragnarok tested by Stereophile?  I would think it does.

[esldude]

Oh no! Such confusion. 

 

Bridging amps.  Let's use some hypothetical examples.

 

100 wpc into 8 ohms in a stereo amp.  28 volts and 3.5 amps (please allow me to drop extra decimals).

200 wpc into 4 ohms in a stereo amp.  28 volts and 7 amps. 

 

Now if we bridge the amps the way is being done with the Vidar our two channel amp becomes mono.

400 wpc into 8 ohms.  56 volts and 7 amps.

 

How did we get 56 volts?  Because each channel is being run in opposite phase so you really have +28 volts and -28 volts for a 56 volt difference.  You will need twice the amps and it results in 4 times the power.   You don't gain any increase in amperage this way because one channel is feeding 7 amps in one direction and the other must take it in the opposite direction.  If 7 amps is the most your amps can put out then it has become a limiting factor.

 

So you'll get a mono bridged rating of 400 watts into 8 ohms.  Into 4 ohms what will you get?  As 4 ohms will require more amperage than can be supplied at 56 volts the power output will not be able to double. 

 

Does that mean 4 ohms is verboten and can't be used?  No.  The amp has become current limited and into 4 ohms would manage 7 amps which would give only 200 watts.  So obviously bridging the amp for 4 ohm speakers doesn't gain you anything versus just using a stereo amp for the same load and output power.  Though if you did use a mono amp with 4 ohm speakers requiring no more than 200 watts it would work fine.  Bridging does increase the output impedance so damping of the bass might be negatively effected.  Reactive aspects of speaker loading also might require amperage not available and alter the sound in some ways too. 

 

BTW, bridging like this for electrostats is often a very good thing.  ESL's don't often need extreme current.  What they need is higher voltage. 

 

There are other ways to combine two channels that increase current capability without increasing voltage.  Usually not done with SS amps.  When you bridge some transformer coupled tubes amps that is usually what you are doing.

[esldude]

2 hours ago, Speed Racer said:

Vidar is not bridged when acting as a monoblock amp.

I am pretty sure it is.

 

If you feed only single ended inputs for stereo, guess what happens if you have balanced XLR input for mono.  The two channels are out of phase.  You connect up the two positive posts and voila' you have a bridged amp.

 

From the owner's manual:

 

 

When using Vidar as a mono amp, connect your speaker to the two

top red terminals only. Right is “+”, and left is “-”.

 

Please note both of these are active outputs—do

not ground one of them!

 

XLR Input.

Connect a balanced source (usually

a preamp) using a high-quality XLR cable. It’s

important that the preamp is truly balanced for

maximum power output.

 

 

The following is from Schiit's FAQ:

 

 

Wait. How does the monoblock mode work?

It’s not a mode. It’s just the natural result of driving each channel with one half of a truly balanced signal. That’s why you need a preamp with balanced outputs. And that means REAL balanced outputs, not just XLR connectors.

 

There’s a difference between XLR and balanced?

There can be. If you aren’t using Freya, ask your preamp manufacturer if it has true balanced, differential outputs. If it does, you can easily run two Vidars as monoblocks. Or more, if you want to use them for home theater or something. 

 

Same for 4 ohms when running monoblocks?

Yes. Again, Vidar will probably work fine at sane volumes, but at higher output, you may trigger the protection. 

 

So what’s a Vidar?

Vidar, in Norse Mythology, is said to be the strongest of the gods after Thor. And no, this isn’t a hint that there’s an even more powerful “Thor” amp coming—we just don’t want the copyright battle on that one. Vidar is one of the newer generation of gods that survive Ragnarok. Yep. I know. We’re here all day.

[esldude]

15 minutes ago, Speed Racer said:

I am sure it is not!

 

Here is a quote from Jason Stoddard, the designer, from another forum:

 

"Sorry to disagree, but it is not [bridged]."

 

"Bridging is the process of using the output of one side of the amp to drive the inverting input (feedback point) of the other side of the amp, in a typical Lin-topology amp. This is technically "in series" and can have sonic consequences that are not ideal."

 

"Vidar is not a Lin topology. When used as a monoblock, it is simply amplifying a differential signal. It is not bridged."
 

 

 

You give a tremendous example of why I dislike Jason's approach to promoting his products.  Why I asked in this thread why this amp was disruptive when nothing about it is other than the PR.   Here they are pretending to be doing something different when it amounts to a tiny point of difference about how some others do bridging.  Some do the same exact thing Schiit does.  It can be done both ways, and it amounts to the same difference.   The other way is one needed if you only have single ended inputs.   The Schiit way is also the cheapest way to do it with RCA or XLR, but not both.  Nothing wrong with that in a bang for buck amp like the Vidar. Its smart.  It isn't revolutionary or disruptive.

 

A bridged amp has opposite phase in the two channels.  You tap the positive side outputs only and the differential voltage is doubled.  Current capabilities halved.  Power output quadruples.  Exactly what Schiit is doing.

 

You can invert one channel to feed the other which some people do.  You can simply used balanced input which can be switched from feeding one channel to feeding two and ignore the other input (usually with a switch to swap connections), or you can feed a transformer which can be switched for the different inputs needed.  There are other ways, but those are most common.   You could call it a differential amp instead of bridged as he does.  The way it works for current and voltage and power is all the same.  Just like a Sumo Andromeda from the 1980's. 

 

 

[esldude]

44 minutes ago, Speed Racer said:

This "no it all" type of response is what turns me off to this site. Do you really think you know more about what is and isn't a bridged amp than Jason does?

 

How many amps have you designed and successfully sold? What are your credentials that allow you to admonish Jason like this?

I do in fact know enough to understand what the Vidar is going in regards to stereo vs mono.  That is all I need to know in this case.  What it really and truly in the physical world is doing is as I describe it and credentials or no, Jason Stoddard's pronouncements or no, that is what it will do.  In fact I doubt Jason would have a big point of disagreement with me.  The only real point would be terminology of balanced differential rather than bridged.  The functioning for evaluating power and speaker loads is the same.

 

You don't understand it or you wouldn't be arguing with me.  That is okay, I can overlook it.  But don't start off on some big rant about Jason Stoddard vs me.  This is not what people wish to read here.  I've posted info that is useful to understanding what the Vidar will do in regard to different speakers loads and mono vs stereo.  The info is solid and agrees with literature from Schiit.  I'll leave it at that.  If you are not convinced....too bad.  The part about whether the Vidar is bridged or balanced differential is trivial and has no bearing on the issue of its power. 

[esldude]

1 hour ago, Don Hills said:

A bridge circuit takes an input signal (single ended or differential) and produces two equal and opposite output signals. A "bridged" amplifier has this bridge circuit at its input. In the case of the Vidar, the differential output stage of the preamp is used as the bridge circuit instead, hence Schiit's insistence that the preamp output be truly differential. If it's not, and the two wires of the preamp output aren't equal (and opposite) in level, the two channels of the Vidar will not be driven equally.

 

So when Jason says the Vidar is not bridged in mono mode, he's strictly correct, there's no bridging circuit to switch in. Each channel remains separate. In theory, you could take the cable from the differential output of the preamp, cut off the XLR plug at the amp end, and wire the two signal lines to the two RCA inputs and get the same result as using the XLR input.  Of course, in practice the RCA and XLR inputs may differ in detail.

Which is why it brought to mind the balanced differential Sumo Andromeda which used to work just that way.  You needed xlr to RCA adapters from a balanced preamp to get best results.

 

 

 

[esldude]

1 hour ago, unbalanced output said:

 

Not quite sure of what kind of speakers benefit of high power at high impedances.

 

 

Electrostats.  Most power as you noted is in the bass.  Certainly 500 hz and below covers most power requirements in music. Electrostats typically have a high impedance in the bass which drops at higher frequencies.  They need a high power amp mainly for the extra voltage output.  They are not requiring much current at lower frequencies. 

 

 

Here is a Quad ESL63.  It reaches about 32 ohms there around 100 hz.  A Vidar probably would do great running in mono on those.  Now here is a Soundlab A1

 

 

There is enough reactance there you can't have a totally wimpy amp on current, but it needs voltage more than current.

 

Sanders made an amp designed for ESL loads and reactances.  One version was 1000 wpc at 8 ohms.  That meant it could put out a signal of 90 volts.  Which meant it was able to push out power into the lower frequencies of ESL speakers.

 

Now in the case of Quad they limited input to 40 volts peak or 28 v rms to prevent arcing over the speaker.  If you did, it shorted the amp inputs to prevent speaker damage.  I seem to recall the Quad 57 shouldn't receive more than 33 volts peak to prevent damage.  I fixed some warning LEDs on the 57s I had to alert you when you neared 28 volts. 

[esldude]

Well Martin Logans CLX or the Soundlabs come to mind.  You want a full range model, not hybrids with cone woofers. 

 

What is the current Quad is it the 2912?

 

Audiostatic is one I would like to see reviewed.  Audiostatics are not so highly expensive as the other full range ESL panels have become.   They ship worldwide direct to the customer which seems like a good plan in the internet age.  So if they would agree to let you review some that would be my choice. We need affordable full range ESLs. 

http://www.audiostatic.com/models.html

 

There is also the King sound ESLs.  They are less expensive than some though I don't know much about them.

http://www.kingsaudio.com.hk/speaker.html

 

Any other full range ESL panels out there?

 

 

 

 

[esldude]

Maggies are all well and good and are panels.  They aren't electrostatic.  Funny, as my first panel speakers were some second hand MG2i's.  The owner also thought they were electrostatic.

 

The Maggies are in fact pretty close to pure resistance.  So while they are 4 ohms, and not highly efficient they require less of an amp current-wise than say a nominal 6 ohm ported box speaker is likely to ask of the amp. Due to the reactance  from the use of a port.  Not downing the Maggies or using them with the Vidar.  Just pointing out they don't fulfil Chris' interest in auditioning some electrostats with the Vidar because the Vidar would appear in mono mode to be better than usual with the unusual load of a full range ESL.

 

Maggie 3.6 impedance plot. Notice how below 1000 hz the phase is 0 degrees.  No reactance.  And the impedance is nearly a straight line between about 4.5 and 5 ohms. 

[esldude]

RAM labs tubes are consistently good.

[esldude]

3 hours ago, crenca said:

 

Thanks for posting this I was not aware of them.  However, they don't appear (unless I missed it) to offer a 6SN7 variant...

You are right.  I was thinking a 6922 would work for a 6SN7, but my tube skills have atrophied and I was thinking of the 5692. 

 

These guys have some 6SN7 tubes.

 

https://www.tubedepot.com/t/tubes/preamp-tubes/6sn7

 

Those RCA's would be nice if NOS.

 

In case you don't know of these guys, but most do, they have all NOS stock. 

https://tubeworldexpress.com/

 

 

 

[esldude]

4 hours ago, unbalanced output said:

I know this link. The error is they only think about attenuation and don't consider the effect of the cable load on the amp (like of every amp would be perfect). If they what they write was true, you could have one cable 20m longer than the other and just slightly compensate the balance. Reality is, that's not true. I started with a generic cable with recommended gauge and length, then switched to the NAC A5 - according to these guys the attenuation should be the same, reality is that the A5 was louder by more than a dozen dBs. I was astonished.

 

Ps. I remember Naim don't put a constraint on the maximum cable throw, their concern is if lengths are too short or different. If remember correctly, they say that for more than 20m one may need their more powerful models.

Frankly I do not believe this.  You changed cables and volume changed by more than 12 db?  How did you measure this volume change? Naim NAC A5 are 6 gauge cable, but more than 12 db doesn't make sense.

 

[esldude]

46 minutes ago, JSeymour said:

I had a similar experience.  I had a pair of Signal Cable Ultras (6 ft & 15 ft) that I used with a Class D Audio SDS-470 with no issues.  I first replaced it with a single Vidar with no issues.  But when I went to 2 Vidars, I could not get either of them to work with the 15 ft. cable.  So I temporarily rigged up bare wire and it worked. 

 

I used a Fluke meter at work and the 15 ft cable had a very small amount of leakage between the 2 wires.  Not enough to be a short, but enough to register on the meter.  My other cable (6 ft) showed nothing.  Nick, Schiit tech support, through a series of emails, said that the Vidar in Mono will show greater sensitivity to cable issues due to the increased current capability.  Nick was very helpful and patient.  In the meantime, I have put in another 6 ft. Signal Cable Ultra with no problems.  I did not detect any damage in the longer cable, but it went along the front of my fireplace. So I suspect I dropped a log on it at some point.

 

If you can, test the cables with a meter.  You may find the results very interesting.

Not disputing what you are saying, but the Mono should not have increased current capability. 

 

Also makes me wonder if the Vidar has the same monitoring circuit that alters the bias which caused trouble when JA tested their amps and suffered growing distortion at sustained signal levels.  Seems this same circuit may be getting tripped too easily if the cable load is not exactly what the Vidar needs to see.  Schiit should probably set out some parameters for cable if that is the case, though with the variability in speakers that would be difficult. 

[esldude]

3 hours ago, The Computer Audiophile said:

@esldude Any idea what might be happening with my Vidar cable experience?

Typically an audio amp will have parallel R and C with the output to help keep it stable.  I wonder if using the microprocessor to monitor things in the Vidar as was done in the Ragnarok Schiit didn't either use minimal or perhaps no output network.  Therefore you may need a minimum amount of parallel capacitance in the cable you use.  15 ft vs 8 ft plus the MIT Z was wound in a way to have rather high parallel capacitance.  So it fixed your output stability problem.  Without knowing the details though that can only be a guess. 

 

If you want to test it out, get a 100 nanofarad capacitor place it across the leads either at the amp or speaker.  Connect up the 8 ft cables and see if the amp is suddenly happy. 

[esldude]

3 hours ago, unbalanced output said:

 

I estimated by memory from the Foobar volume control, it is by no means a precise figure but I'm willing to test it sometimes when I buy a calibrated mic (will take some time - my budget is compromised for the next three months already). True, I struggle to understand why the difference too but the fact is that it was a lot louder after the change. It was a big surprise to me. I used to be a cable disbeliever until then! Resistive attenuation alone cannot explain the difference, so I humbly speculate it has to do with a somewhat high capacitance of the generic cables - I always read that Naims abhor high capacitance cables. 

 

@The Computer Audiophile

If you have a smartphone, download a SLM app.  Send pink noise across.  See what the difference is.  The smartphone won't respond much to frequencies below 200 hz, but would do to show a difference in volume this large. 

[esldude]

3 hours ago, JSeymour said:

 

The increased current capability is the answer I got from Schiit.  If you are now only driving one channel in mono, versus driving two channels, I would call that increased current capability.  The specs in stereo are 100 watts @ 8 ohms, in mono they are 400 watts @ 8 ohms. 

Read some of my posts early in this thread.  In mono mode Vidar is running each channel out of phase.  So you get twice the voltage, but current capability is the same as one channel in stereo mode.  So no increased current capability.  Twice the voltage will give you 4 times the power up until you run out of current.  That is why the amp is not rated as highly with 4 ohm speakers in mono mode.  You'll run out of current at high power and lower impedances.  You likely will have no problem with 4 ohm speakers, but you'll get no more power from them than you will in stereo mode. 

[esldude]

39 minutes ago, esldude said:

 

If you want to test it out, get a 100 nanofarad capacitor place it across the leads either at the amp or speaker.  Connect up the 8 ft cables and see if the amp is suddenly happy. 

Upon further consideration without knowing what Schiit is doing scratch the above idea.

 

It would be safer to put a 1 kohm resistor across the speaker or amplifier leads.  Use at least a 2 watt one.  For permanent installation I would want something more like 10 watts. The idea is the amp may be oscillating at some high frequency like megahertz frequencies and shutting down.  The resistor will damp that oscillation some. 

 

Now some small capacitance value would tune that out of the amp, but a shot in the dark value might not be a good idea.  Would be nice if Jason would give us an idea about how they have the output setup.

[esldude]

8 hours ago, jmsent said:

But if this is indeed what's going on, and those cables are operating normally (not shorted) then this raises serious red flags about the design of these amps. Knowing who these engineers are, I find that  hard to believe.  Good grief, the industry has  only been designing solid state amplification since the mid 1960's. Thousands of products  have been made that are unconditionally stable into virtually any load. Something this modern and with this engineering pedigree shouldn't be acting like this. Either you've got a bad amp or Schiit needs to go back to the drawing boards.

That may be going to far.  Spectral amps due to wide bandwidth also had to use interconnects and speaker cables with a bit of care.  Fine design nonetheless.  The Vidar could be as well.  It might be good if Schiit were a little more informative about which situations or parameters might accidentally trigger a shutdown. 

[esldude]

12 minutes ago, The Computer Audiophile said:

I just tried a third pair of high end speaker cables and had the same issue. All the cables that have "failed" to work with the Vidar have been 8 feet long and connected to my TAD CR1 speakers. 

 

Much more to come. 

Could you list all cables? Might be possible to see if it's capacitance or inductance causing the problem.

[esldude]

Guess it is a moot point for now.  The Kimber has plenty of parallel capacitance so I doubt that was it.  The short cables make have too little series inductance and are allowing an amp oscillation at an ultrasonic frequency. 

 

I was going to suggest putting a .1 microfarad capacitor in series with a 5 to 10 ohm resistor.  Then putting that across the amp leads.  That might damp out an ultrasonic resonance without changing much else. 

[esldude]

8 minutes ago, Zorlac said:

Maybe I should look into something like the CMX-2 from Emotiva to remove DC Offset?

Why don't you put a multi-meter set to DC volts on them first and see if you have any DC offset.  Just set it to DC and read across the two binding posts of the speaker.

 

While you are at it, switch it to AC volts with no signal going through the system and see what AC volts you read.  How useful that is will depend a bit on your meter, but would be helpful maybe to see how much the buzz level is in your setup. 

 

The link above btw is referring to DC offset at the amplifier output to the speakers.  The Emotiva unit would remove it from your AC power lines.  Two different things. 

 

Finally why not plug your amps directly into the wall without the surge protectors to see if that makes it diminish the buzz or go away. 

[esldude]

1 hour ago, Zorlac said:

Okay..here are the multi-meter results:

 

AC

Left speaker = 0.003V

Right speaker = 0.003V

 

DC

Left speaker = 5mV

Right speaker = 0mV

 

I hear the buzz in both speakers when I put my ear up to the cones.

Can anyone tell me what these readings mean and what they should be?

 

 

The AC value is a bit high. 

 

For instance if you had 90 db speakers at 8 ohms, then 2.83 volts would output 90 db.  The ratio of your AC value to that is about -59db.  So you would expect some AC signal level at about 31 db.  Which would be audible.  Of course it may not be the amp's fault.  It will amplify any noise fed to it from upstream.  One way or another however I would want lower AC output than the .3 mV you are showing.

 

This isn't out of line with hearing the buzz when your ears are close to the speaker.  If you turn everything off, and then turn on just the amp do you get a lower value?  Does the buzz go away?  Also might help to know your signal path.  What source, to what DAC and/or preamp before going into the amp.  And what speakers are you using.

 

The DC offset is fine.  Rule of thumb is 50 mV or less.  Some say 25 mV or less.  Either way your DC value isn't a worry. 

[esldude]

6 hours ago, schiit said:

Hey all,

 

Okay, so I'm back in the office, and I spent some time with Chris’ problematic Vidars. And yep, there’s something up with them, something that’s triggered by the cable he sent us. Apparently the cable’s combination of capacitance and inductance sets the amps off, even though Vidar has (a) a full Zobel network, and (b) is tested into punishing reactive loads.

 

What’s weird is that we can’t reproduce the problem on any of the Vidars we have currently in stock.

 

Because of this, I dug into Chris’ amps some more:

• Tested the Zobels to make sure they were connected and not fried—they were fine
• Replaced the Zobels with current production parts—no change
• Changed the compensation capacitors to current production parts—and that fixed it

Now, this makes no sense. Changing a like-for-like part shouldn’t make any difference. Unfortunately, I didn’t keep the comp caps I took out of Chris’ amps to measure (and you can’t measure them in circuit, they’re tiny, low-value 0805 SMD parts.)

 

So what happened? I’m not completely sure. Perhaps some amps got the wrong comp caps in them, even though they are surface-mount and robotically placed, and you’d think that would be very unlikely. The parts are unmarked, and most small SMD caps look the same, so they can’t be identified visually.

 

Because of this, we’re adding a new stability test to the Vidar production process, using an RLC network that mimics the cable Chris sent us. This is in addition to the standard process which includes automated instrumented test, load testing, 1 day burn-in, and a listening test. We’ll also notify Vidar owners in the same batch as Chris’ amps that if they are experiencing operational oddities, we’ll swap them out at our cost.

 

But, in any case, if you have Vidars that are behaving oddly, contact us and we’ll get you taken care of. Sorry for the trouble!

 

(Oh, and Chris, we'll get the cables back to you today.)

 

All the best,

 

Jason Stoddard

Co-Founder

Schiit Audio

Kudos. 

 

This is how a company should handle problems for their customers when they crop up.  So many would be in denial and obsfuscation instead of handling the problem.