AMB Laboratories DIY Audio

[Iniamyen]

I'd really like to understand your position that the 300mW is unrealistic. Please correct me if any of my calculations are incorrect:

300mW into 33R = 3.1464 Vrms
3.1464 Vrms = 8.81 Vpp

How is this not possible with a 10.XXX-Volt battery? Do you expect diode drops, etc... to exceed 2V?

[NwAvGuy]

This is the DIY community, not some company trying to sell products based on false claims. So why the Ralph Nader approach? No trophies for you here...
Starting off with "Hey guys, I know a thing or two about audio, how can I help?" would have been great.

It would be a waste of your talents if you became persona non grata in the community because of the attitude.
Perhaps it's not too late to start over?

I appreciate the feedback. I get products and I review them. I try to just tell it like it is. I pointed out several good things about AMB and the Mini3 in my review. But, no, I don't sugar coat things and perhaps I should try a softer approach.

With NuForce I did contact them before publishing my review to find out if there was something wrong with the uDAC-2. They explained it was working as they designed it so I went ahead and published the review. When I disputed some of their specs, their response was to admit their specs were wrong and they dramatically changed a few of their published numbers. I expected something similar here.

I posted a link to the Mini3 review here first. It didn't get approved for quite a while and I wasn't even aware of this thread having started. If you go back to where I joined this thread I was trying to be as civil as possible and simply address factual concerns like the grounding of my equipment, etc. But the posts from nearly everyone else from, before I even started posting, were negative towards me and the review. So I started from a defensive position by default.

Exactly one person suggested I work with AMB to improve the Mini3 design, and I said I'd be open to that. But that, obviously, hasn't gone anywhere as nobody at AMB even admits there are problems. I really didn't expect the designer of the Mini3 (if that's who "AMB" here is) to try and defend numbers there's no rational way for anyone to defend.

I think that's a critical point to understand. A lot of people build AMB designs on the assumption they're carefully engineered proven designs. There are lots of impressive numbers published in the specifications section for most AMB projects. I assumed, this being a DIY site, those numbers were reasonably accurate. I was really surprised to find out how many of them are not just wrong, but wrong by huge margins. I was even more surprised to have the AMB guys try to defend the obviously erroneous numbers.

Isn't anyone else here bothered by that? Am I really the bad guy for pointing out some of the AMB numbers are seriously wrong and trying to show exactly how they're wrong when someone tries to defend them against all odds?

If Ford says the Mustang can go from 0-60 in 6.0 seconds, and the reality is when properly tested it's really 18 seconds--slower than Prius--isn't that something people would like to know? Why is Ford the good guy in this scenario?

[NwAvGuy]

I'd really like to understand your position that the 300mW is unrealistic. Please correct me if any of my calculations are incorrect:

300mW into 33R = 3.1464 Vrms
3.1464 Vrms = 8.81 Vpp

How is this not possible with a 10.XXX-Volt battery? Do you expect diode drops, etc... to exceed 2V?

Your numbers above are correct. Here's the real deal:

The biggest drop is about 1.7 volts across the 6.2 ohm output resistor. Add another 1.1 volts or so for the AD8397 and the 1N4007 diode. And the battery voltage, per multiple datasheets, is really 9.5 volts best case under load right off the charger. So the real math:

9.5 - 1.1 - 1.7 = 6.7 V p-p = 2.37 V RMS = 170 mW RMS (for about 10 minutes until the battery drops to 9 volts)

But that's hard clipping with just one channel driven. If you want just 1% THD it's significantly less. And the battery voltage is going to quickly sink.

Now, if you want to drive 2 channels into 33 ohms into clipping, the OPA690 falls on its face and yanks the power down still further. The above 170 mW in 33 ohms is 101 mA per channel or 202 mA total from the OPA690. It's rated at 160 mA - 190 mA.

If that's not clear enough, please go back in the thread. I've run the numbers backwards and forwards--literally. I've also explained exactly where my 98 mW came from and why.

[bcg27]

This is the DIY community, not some company trying to sell products based on false claims. So why the Ralph Nader approach? No trophies for you here...
Starting off with "Hey guys, I know a thing or two about audio, how can I help?" would have been great.

It would be a waste of your talents if you became persona non grata in the community because of the attitude.
Perhaps it's not too late to start over?

I appreciate the feedback. I get products and I review them. I try to just tell it like it is. I pointed out several good things about AMB and the Mini3 in my review. But, no, I don't sugar coat things and perhaps I should try a softer approach.

With NuForce I did contact them before publishing my review to find out if there was something wrong with the uDAC-2. They explained it was working as they designed it so I went ahead and published the review. When I disputed some of their specs, their response was to admit their specs were wrong and they dramatically changed a few of their published numbers. I expected something similar here.

I posted a link to the Mini3 review here first. It didn't get approved for quite a while and I wasn't even aware of this thread having started. If you go back to where I joined this thread I was trying to be as civil as possible and simply address factual concerns like the grounding of my equipment, etc. But the posts from nearly everyone else from, before I even started posting, were negative towards me and the review. So I started from a defensive position by default.

Exactly one person suggested I work with AMB to improve the Mini3 design, and I said I'd be open to that. But that, obviously, hasn't gone anywhere as nobody at AMB even admits there are problems. I really didn't expect the designer of the Mini3 (if that's who "AMB" here is) to try and defend numbers there's no rational way for anyone to defend.

I think that's a critical point to understand. A lot of people build AMB designs on the assumption they're carefully engineered proven designs. There are lots of impressive numbers published in the specifications section for most AMB projects. I assumed, this being a DIY site, those numbers were reasonably accurate. I was really surprised to find out how many of them are not just wrong, but wrong by huge margins. I was even more surprised to have the AMB guys try to defend the obviously erroneous numbers.

Isn't anyone else here bothered by that? Am I really the bad guy for pointing out some of the AMB numbers are seriously wrong and trying to show exactly how they're wrong when someone tries to defend them against all odds?

If Ford says the Mustang can go from 0-60 in 6.0 seconds, and the reality is when properly tested it's really 18 seconds--slower than Prius--isn't that something people would like to know? Why is Ford the good guy in this scenario?

I think most people appreciate independent measurements. I think the issue here is that we have one set of measurements coming from the designer of the amp, who from everything I have read has a ton of experience in audio and has produced a lot of well received designs. Then we have another set of measurements coming from you, who is at least to me completely unknown in the audio world. Who would you believe? It also doesn't help that, at least from my point of view, you came in with a very aggressive and condescending tone. I honestly don't know much about testing audio equipment as I am one of your EEs that didn't spend much time on analog circuit design outside of required intro courses, but I would need to see another set of results that confirm what you are seeing before I completely trusted them. I bet a lot of people here feel the same way.

I have not personally built the mini3, but I have build the B22 and gamma2 and they both sound really good to my ears. Like you say in your blog you don't necessarily hear some of the bad measuring characteristics, but I would be interested in seeing your measurement results for those devices if you ever do measure them.

Also as far as I know there are no 'AMB guys,' AMB is one guy who sometimes teams up with other people to design these products.

[Iniamyen]

The biggest drop is about 1.7 volts across the 6.2 ohm output resistor. Add another 1.1 volts or so for the AD8397 and the 1N4007 diode. And the battery voltage, per multiple datasheets, is really 9.5 volts best case under load right off the charger. So the real math:

9.5 - 1.1 - 1.7 = 6.7 V p-p = 2.37 V RMS = 170 mW RMS (for about 10 minutes until the battery drops to 9 volts)

But that's hard clipping with just one channel driven. If you want just 1% THD it's significantly less. And the battery voltage is going to quickly sink.

Now, if you want to drive 2 channels into 33 ohms into clipping, the OPA690 falls on its face and yanks the power down still further. The above 170 mW in 33 ohms is 101 mA per channel or 202 mA total from the OPA690. It's rated at 160 mA - 190 mA.

If that's not clear enough, please go back in the thread. I've run the numbers backwards and forwards--literally. I've also explained exactly where my 98 mW came from and why.

I guess that you're using a calculation to try to show why someone's measurement must be a lie. I guess most people would be inclined to trust the measurement, given that it's probably closer to reality. If you wanted data to support your argument, you could measure Vpp into 33R like the original test and see what you get.

[NwAvGuy]

I think the issue here is that we have one set of measurements coming from the designer of the amp, who from everything I have read has a ton of experience in audio and has produced a lot of well received designs. Then we have another set of measurements coming from you, who is at least to me completely unknown in the audio world.

I agree that's one of the bigger issues here. And I welcome another valid independent set of measurements. With NuForce it was easier as they verified, and agreed with, several of my more critical measurements. I don't think they'd ever bothered to even make some of the measurements before I came along. And they admitted as much to Stereophile when their CDP-8 tested poorly by John Atkinson.

But once they fixed their misleading specs, NuForce did what a lot of the audiophile community does. When faced with bad measurements they just turn the argument into "we care about good sound more than measurements." And that's fine. If someone wants to trust their ears above all else, subjective bias and all, that's their choice. If certain gear puts a smile on their face, that's what matters most.

But I do have a problem with anyone publishing specs that are grossly misleading. Some people make buying decisions based on specs--just like the guy shopping for a new car. And if someone's publishing bad specs that others are using to make decisions I think that needs to be brought into the open.

You can't drive into your local AMB Shop and audition the projects here. So in some ways the specs are more critical than buying a car. You have to largely trust online information to make your decisions. I just think that information should be reasonably accurate and not misleading.

WIth DIY it's even more important, because unlike a NuForce DAC, you usually can't just return it once you invest all the time and money to build it. I think most people contemplating DIY projects would like to base their choice on accurate information.

[NwAvGuy]

I guess that you're using a calculation to try to show why someone's measurement must be a lie. I guess most people would be inclined to trust the measurement, given that it's probably closer to reality. If you wanted data to support your argument, you could measure Vpp into 33R like the original test and see what you get.

EDIT: We've come full circle here. All the calculations were only to explain my measurements and try to demonstrate what's even possible for the purposes of evaluating AMB's claimed measurements.

I did measure it into 33 ohms. I quote from my review:

"The Mini3 only managed 1.8 V RMS into 33 ohms or 98 mW at 1% THD. That’s only 1/3 of what AMB claims."

[Beftus]

You can't drive into your local AMB Shop and audition the projects here. So in some ways the specs are more critical than buying a car. You have to largely trust online information to make your decisions. I just think that information should be reasonably accurate and not misleading.

WIth DIY it's even more important, because unlike a NuForce DAC, you usually can't just return it once you invest all the time and money to build it. I think most people contemplating DIY projects would like to base their choice on accurate information.

The desire to do DIY is sometimes bigger than the abilty to finish the project satisfactory. Therefore I think most people approach DIY differently than when buying something commercially produced in a shop. I know I do.
After having read a lot about the Mini3 I made myself one expecting to end up with a very good sounding (and loud!) headphone amp. While I know that online information should often be taken with a grain of salt, I don't feel misled by AMB (if he even did that). The Mini3 didn't disappoint me, and still doesn't. I use it almost on a daily basis. Specs don't mean much to me, I judge based on what I hear.

BTW If you ever get around to making an improved Mini3 gimme a call. I'd be more than happy to test it.

[egallego]

I appreciate the feedback. I get products and I review them. I try to just tell it like it is. I pointed out several good things about AMB and the Mini3 in my review. But, no, I don't sugar coat things and perhaps I should try a softer approach.

Surely everybody here appreciates your technical feedback. The failure was with your tone. Maybe AMB or you did a mistake with the measurements, or they are simply not comparable due to different testing conditions.

I posted a link to the Mini3 review here first. It didn't get approved for quite a while and I wasn't even aware of this thread having started. If you go back to where I joined this thread I was trying to be as civil as possible and simply address factual concerns like the grounding of my equipment, etc. But the posts from nearly everyone else from, before I even started posting, were negative towards me and the review. So I started from a defensive position by default.

It didn't help you used a different design that the one published in AMB site.

Isn't anyone else here bothered by that? Am I really the bad guy for pointing out some of the AMB numbers are seriously wrong and trying to show exactly how they're wrong when someone tries to defend them against all odds?

I'm just curious to see what is the outcome. Either AMB made a mistake, you did the mistake or simply the testing is not comparable. In either case I don't see the problem. Anyone who has designed something will make mistakes. I'm very sure that nobody did it intentionally. Indeed, I guess everyone is very grateful to AMB given the enormous help and contributions he has done to the community. When you get the design "for free" you can't really complain in such manners, instead, you should contribute and help the design.

[NwAvGuy]

It didn't help you used a different design that the one published in AMB site.

Even AMB has admitted the Mini3 I tested is electrically identical to the published design when running from battery. So this is just a straw man that keeps re-appearing.

instead, you should contribute and help the design.

I've already offered that. But I can't help with problems AMB insists don't exist. It's unfortunate.