Densen amp

Administrator
Joined 2007
Paid Member
Mooly, old chap, how do you do? Are you working on something?

I am fine thank you, hope you are too.
No serious audio projects on the go at the moment... still more than happy with the "Mooly amp". It still sees of the competition.... what was it you said at the time... I always remember, something like "It's so easy isn't it" :)

I have concluded you probably won't change peoples opinion on here, the same topologies appear, in one form or another, with sound quality often in inverse proportion to the complexity of the design. I accept too that perceived "sound" is so very personal, yet I'm sure many haven't really seen (or heard) lol the light when it comes to audio reproduction.

Interesting non the less, seeing what everyone is up to.

A little fun project/modding exercise here,
http://www.diyaudio.com/forums/soli...l-ra820a-use-hexfetss-amongst-other-mods.html

Apologies to the original poster too, perhaps I can add this as some practical experimentation on excluding the outputs from the global feedback as per Lumanauw's diagram.

http://www.diyaudio.com/forums/soli...-hexfet-poweramp-sonic-benefits-approach.html
 
Well... I was "the other half" of the experiment Kurt von Kubik has described above. I don't remember the excact current in the SE, but I do ermember that we vere using to-126 transistors to dissipate the heat. My guess is that we had the SE running at approx 70-100mA. And the difference between SE and PP was huge. Running the SE at 200mA may change a little, but not that much.

We did a lot of testing some years ago, to decide what to use...: SE, PP or op-amp. And every single test pointed directly towards the PP version.

Anyway. This thread in NOT about SE, PP or op-amp buffers. It's about some Densen amplifier.

You are right. Back to the actually topic. About the weblink
Densen Audio Technologies
Densen’s unique "DMCD" (Densen's Mass Current Distribution) technology (patend pending) was mentioned. Also about
Densen Audio Technologies
Are there already a patent document?
Please note, through the suppression of the exact URL and the transformation to the headline of each webaddress-URL about this forum, you think, this two links are identical. But this isn't so.
 
Last edited:
I do not know about this patent. But looking at the schematic shown, I do not see anything that is worth a patent.
Wonder what the patent is all about :confused:

What I do know is, that the B-110 is based on 1 pair of 2SC2922/2SA1216. This seems very OK for an amplifier rated at 60 watt@8ohm and 120watt@4ohm. Only problem is, that it has no heatsinks, and therefor runs hot. During the lawsuit, the independant engineer testing the B-110, proved that the B-110 was capable of less than 12 watts pr channel @ 4 ohms, due to overheating. That is less than 1/10 of the rated power. Fitting a heat sink into the B-110 would in my opinion make it an acceptable amp in it's price range. After all you do get a NFB design, relay controlled attenuator in a nice design.

However.... The big issue if you ask me, is that when going from the B-110, that seems OK in terms of output devices, to the B-350 capable of around 300 watts, Densen still use the same 1 pair of 2SC2922/2SA1216. That seems to be a very weak solutions for 300 watts, and as far as I know, way above the reccomendations from Sanken.
When you also take into account, that B-350 is supposed to be some sort of High End Statement product, this seems even worse.
 
Last edited:
The DMCD patent I think is unknown to everybody exept Densen themselves I think.
My gues about what it actually is meant to be, is the distribution of electrolytic capacitors to the sites throughout the amplifier, where the power is needed, rather than having your reservoir placed centrally in fewer bigger electrolytics.
That´s what I´ve hard about it so far. IRL this means one central PSU with smoothing caps, which often in Densen gear are very large, even in line level gear, and then a series of smaller electrolytic caps placed near the active components.
Maybe they found out a certain relationship between large an small caps, which they hen filed a patent for, I don´t know.
 
Last edited:
A1216/C2922 are 200W devices, so a single pair does 400W.
Pd of a power device is the max dissipation level it can handle and does not directly translate to output power, imo.
The power levels Densen issued for a single pair output stage suggests a fancy protection circuitry, aka a dissipation limiter.

The 225W-8Ohm Kenwood M2A had only 4 Sanken DAT devices per side in a low/high switched power configuration.
DAT devices are two matched power transistors in parallel on 1 substrate, aka Dual-Audio-Transistor, with BE-C-EB pins.
You can find the M2a schematic at audio-circuit.dk of our diyA knight in shiny armour Jan Dupont.


I'm sure what you mean. But you will NOT be able to dissipate 200 watts pr Sanken device. This assumes that you can keep the temperature at 25 degree C, which may be hard when dissipation 200 watts ;)

At 100 degree C the deratin curve show maximum 80 watts. In real life applications, I would go for something like 50-60 watt pr device! And not a single watt more!
 
I do not know about this patent. But looking at the schematic shown, I do not see anything that is worth a patent.
Wonder what the patent is all about :confused:

What I do know is, that the B-110 is based on 1 pair of 2SC2922/2SA1216. This seems very OK for an amplifier rated at 60 watt@8ohm and 120watt@4ohm. Only problem is, that it has no heatsinks, and therefor runs hot. During the lawsuit, the independant engineer testing the B-110, proved that the B-110 was capable of less than 12 watts pr channel @ 4 ohms, due to overheating. That is less than 1/10 of the rated power. Fitting a heat sink into the B-110 would in my opinion make it an acceptable amp in it's price range. After all you do get a NFB design, relay controlled attenuator in a nice design.

However.... The big issue if you ask me, is that when going from the B-110, that seems OK in terms of output devices, to the B-350 capable of around 300 watts, Densen still use the same 1 pair of 2SC2922/2SA1216. That seems to be a very weak solutions for 300 watts, and as far as I know, way above the reccomendations from Sanken.
When you also take into account, that B-350 is supposed to be some sort of High End Statement product, this seems even worse.
Densen mentioned
"the B-350 delivers 125W at 8 ohm, and 250W in 4 ohm"
but he don't mention the conditions of measuring (sine wave, RMS, program power) Therefore I think, the mentioned power output is to interpr. as "PMPO". Please read concerning this:
Peak Music Power Output - Blue Room wiki
 
Don't think you will hear Densen caliming that the rated power is PMPO. During the lawsuit, Densen themself accepted the standard to which the power should be measured. And even so.... They lost it!
According to the standard accepted by Densen, a B-110 is capable of delivering less than 12 watts @ 4 ohm.
 
Don't think you will hear Densen caliming that the rated power is PMPO. During the lawsuit, Densen themself accepted the standard to which the power should be measured. And even so.... They lost it!
According to the standard accepted by Densen, a B-110 is capable of delivering less than 12 watts @ 4 ohm.

That means automaticly relatively low supply voltage at the output buffer stage for high realibility regarded lifetime and stability for idle current.
If all this estimates are right, I would good understand the avoid of schematic releases for service and maintenance.
 
Mmm, the idea for the B-110 likely was that the aluminum extrusion bottom/sides part of the amp case is the heatsink.
 

Attachments

  • Beatme-110.jpg
    Beatme-110.jpg
    100.9 KB · Views: 947
Mmm, the idea for the B-110 likely was that the aluminum extrusion bottom/sides part of the amp case is the heatsink.
the mentioned peak power by the German magazine "Audio" are 75 watt/8 ohms and 155W/2 ohms - look at bottom of
http://www.hifisound.de/oxid/out/oxbaseshop/html/0/test_pdf/DSE-2025000.pdf

Instead 300W/2 ohms there are only 155W, that means high value of internal resistance of the power supply resp. transformer. Typical for transformers and capacitors in this size

For commonly home hifi applications this kind of heatsink creation is a good solution, but it must be ensured that selected quiescent current is 20mA until max 30 mA (i.e. normal class A / B).
Older Naim Audio amp models so as Nait, NAP110 or NAP160 use basicly the same solution for the heatsink.

Can you make a high resolution foto of the driver part from power amp section?

Here inside photo collection from the web:
麗質天生 素顏即美-Densen B-110綜合擴大機 - U-Audio 測試報告
http://www.southaudio.com.tw/Audio AADB01.htm
不愛誇耀的深厚潛質–Densen B-200前級擴大機-U-CAR (B-200)
Audio Video
http://www.gz008.cn/audio/20082043_2.html
Densen lanserer sin kraftigste integrerte - Lyd og Bilde (B-175)
http://image.blog.livedoor.jp/pastel_piano/imgs/2/2/22804b23.jpg
http://www.helloav.co.kr/sub_images/200603170427085.jpg (B 300)
http://www.spl.ru/images/docs/1_791_densen-b175inside.jpg (B 350)
 
Last edited:
Kurt von Kubik said:
from
http://www.diyaudio.com/forums/digital-line-level/138230-dac-project-completed-24.html
Originally Posted by sendler View Post
Wow! That's a lot of heat!
Kurt von Kubik: So it is!
Most DIYérs will be suprised of the heat generated by shunts or class A.
Most manufacturers don´t even have an idea of it

As the matter of fact, I won a lawsuit against Densen Audio Technologies, who claimed some of their amplifiers to be partly class A which they proved not to be at all.
Anyone knowing of the heat probs of class A would know by sight, but Densen really thougt they were right
They weren´t.

Densen Audio Technologies

They do not state the calss A power anylonger though, so at least they became a bit more honest towards ratings.

Are the report of your lawsuit against Densen Audio Technologies on the web?

In danish language I find this:
http://www.hifi4all.dk/forum/forum_posts.asp?TID=69601&PN=1&TPN=4 and english translate about language tools
http://translate.google.de/translat...D=69601&PN=1&TPN=4&sl=da&tl=en&hl=de&ie=UTF-8
But it is hard to understand.
 
Last edited:
Are the report of your lawsuit against Densen Audio Technologies on the web?

In danish language I find this:
HIFI4ALL.DK Forum: Alt for skrap censur her and english translate about language tools
http://translate.google.de/translat...D=69601&PN=1&TPN=4&sl=da&tl=en&hl=de&ie=UTF-8
But it is hard to understand.

1.st of all, I did not file any lawsuit, Densen did, and went of with a very blody nose. They are to pay for the party, and now they have proven evidence that their amplifiers will not stand tall to most of the data stated.

Some of th emost important pages of the actual report is found here:
GratisUpload.dk | Upload dine filer nemt og gratis her

Later on, the head honcho from Densen tried to bribe his way to a better conclusion, but he got busted by the independed engineer, which can be seen here.
GratisUpload.dk | Upload dine filer nemt og gratis her
Sorry it is all in Danish, but have a go with google translate or by bablefish, it might end up with something understandable.

But what can be said about i.e. B110 is that according to IEC 60268 it can deliver 76Watts @ 8Ohms and less than 12 Watts in 4 Ohms.
This measurement was done as the temperature limited power delivery.
You also can do the measurement as a power measurement as a function of distortion. That would mean, that the amplifier should be turned up until the stated THD+N is reached, which is 0,05% (-66dB)for every Densen amplifier, and then power is measured. But that would bring the measurements to even lower values. Just around 1 watt @0,05%THD+N for the flagship amp B350 @ 4 Ohm, so that would not help really.

My point against Densen was originally, that they state i.e. 125W @ 8 Ohm, and 250 watts @ 4 Ohm, which normally will raise the eyebrows of anyone who does understand just a bit of the game. Doubling the power when impedance is halfed, is pretty cool. But if this is done i.e. by an amp, specified as a 100 watt @ 4 Ohm amplifier, the amp will surely deliver either 50 Watts @ 8 Ohms or more, and then you might even exceed your statements, and everybody is happy. That is the way it has been done @ Densen, some of their amps cannot even deliver the specified sine power RMS without running way to hot.

So as this thread started, someone wondered about these amplifier data, and liked to look at the schematics, years ago i did to, because power doubling when impedance halfing, NFB design, 0,05% THD+N, 15-40 watts of class A power in a small nice enclosure with hardly any heatsinks, a nice finish and running cool looks pretty attractive to me, even if I have to pay 7.000€ for it.
Compared to my own amp, which is a "real class A" amplifier producing 100 Watts pr. ch. into 6 Ohms, weighing 100 Kg, running @ 55 dgr. C, using 2 Papst fans forcing 180 m3 of airflow past its massive heatsinks just to keep it that low, featuring 16 250 Watts output devices pr. ch. 4 250 W drivers pr. ch. , massive regulators for all of the voltage amplification, THD+N 0,1% or higher, then data and power consumption like the Densen would be just great.

But the truth was pretty different, they unfortunately proved to be largely overstated, and in addition there were no class A watts what so ever.
 
Last edited:
Thank you for the detailed description and the URLs. Nobody can outsmart physical relationships. Many manufacturers therefore avoid precise details and prefer to use non-specific data in their advertisement documents. So the music lovers get good impression and the product selling is more easy.

Good known example is the PMPO for output power.

If I have develope (for serial production) a normal Class A/B power amp with normal values (arround 30 mA) of idle current through the output devices, I can claim, this is "Class A", because it is "Class A" as long as the current through the loudspeaker does not exceed 30 mA (which is not mentioned in technical details of my advertisement, of course) and as long as I don't mentioned certainly standards (e. g. like International Electrotechnical Commission) based to my claim.

But I cannot claim, that is "pure Class A" because this means, the idle current through the output power devices must have the same value than the maximum possible peak current value through the loudspeakers.

Also confused it is by the output power because it depends on the load (complex character in some cases), the voltage value so as the internal resistance of the power supply and sometimes existing thermal protection to reduction the maximum output current (like Densen)
Naim Audio choice for their thermal protection a more consequent solution: above 60 degrees the thermal switch interrupts the 230VAC lines, since it is in series with the main switch, i. e., the amp device is complete switch off above 60 degrees (better solution as the limiting of output power to 10 % of normally values).
 
Last edited:
Thank you for the detailed description and the URLs. Nobody can outsmart physical relationships. Many manufacturers therefore avoid precise details and prefer to use non-specific data in their advertisement documents. So the music lovers get good impression and the product selling is more easy.

Good known example is the PMPO for output power.

If I have develope (for serial production) a normal Class A/B power amp with normal values (arround 30 mA) of idle current through the output devices, I can claim, this is "Class A", because it is "Class A" as long as the current through the loudspeaker does not exceed 30 mA (which is not mentioned in technical details of my advertisement, of course) and as long as I don't mentioned certainly standards (e. g. like International Electrotechnical Commission) based to my claim.

But I cannot claim, that is "pure Class A" because this means, the idle current through the output power devices must have the same value than the maximum possible peak current value through the loudspeakers.

Also confused it is by the output power because it depends on the load (complex character in some cases), the voltage value so as the internal resistance of the power supply and sometimes existing thermal protection to reduction the maximum output current (like Densen)
Naim Audio choice for their thermal protection a more consequent solution: above 60 degrees the thermal switch interrupts the 230VAC lines, since it is in series with the main switch, i. e., the amp device is complete switch off above 60 degrees (better solution as the limiting of output power to 10 % of normally values).

You are right, outwitting mother nature is a hard task:D.
Maybe even only God almighty is to do so;).
Claiming class A in a poweramplifier, makes no real sense to me if you do not state the power delivered as class A. But if you choose to state that the first few watts are class A, then I think it would be reasonable to expect 2 watts of class A power or more.
In the Densen case the B350 was clamed to deliver 40 Watts class A power and 125/250 Watts in 8/4 Ohms.
There was no impedance stated regarding the class A power, so I´d expect 40 Watts of class A power @ 4 Ohms, but would also settle for at least 40 Watts @ 8 Ohms as a reasonable fulfillment of the statement.
But there wasn´t any at all :eek:.
All other Densen amps were stated 15 watts of class A power, still with no load defined. Of course they also had no class A power at all.
And now for the most funny statement, all line level amplifiers, such as preamplifiers, outputstages in CD players, tuners and so on, were claimed to be 6 Watt class A power:nownow:
If you try to calculate on this for a moment, 6 Watts class A power into the normal 10-20KOhm load, which Densens own pre/integrated amps feature throughout the entire productline, then you´d have to put more than 200V into your preamp :D.
But that´s what was claimed in catalogs and brochures, until he, by the help of the court, realized how faulty and misleading this was, and had to give up his lawsuit. And now these statements are removed from printed material.
 
www.hifisonix.com
Joined 2003
Paid Member
Its easy to have an amp that doubles power as the impedance is halved. Here is how its done:-

Decide what the minimum impedance is that you want to state in your spec. Lets assume here it is 2 Ohms.

Measure the power into 2 Ohms

Now simply halve the power into 4 Ohms

Halve it again into 8 ohms.

The reviewer now reads the data sheet:-

50 Watts into 8 ohms

100 wats into 4 ohms

200 watts into 2 ohms

Conclusion of the reviewer: fantastic power supply. Painstakingly engineered for perfection. Of course, it will exercise 'iron fisted control' over the speakers and the bass.

The reality: Its an 80 watt into 8 Ohms amp, 130 into 4 Ohms ans 200 into 2 Ohms (for a few seconds).
 
One of our designers with a profile in here, has drawn my attention to this thread, and I think it is time to clearify a few issues.

Kurt von Kubik did NOT win any case.

We raised a libel case against him, due to his abusive attacks. It was settled out of court, and with great pleasure we noticed after the case, that he wrote on the forum where he made his attacks, that he had behaved “deeply indecent”.

Regarding the test rapport Kurt von Kubik refers to, with was made by an engineer appointed by the court, it was so faulty that we got Dansk Standard, which is the organisation in Denmark responsible for the standard used in the report, to issue an OFFICIAL statement about the report:

(Full statement here: GratisUpload.dk | Upload dine filer nemt og gratis her )

They concluded:

I syns- og skønsrapport BS99-1273/207 refereres til standarden DS/EN 60268-3, men det må
konstateres, at et flertal af de foretagne målinger er baseret på en fejlfortolkning af denne og andre
standarder. Det har medført at måleresultaterne må betegnes som fejlagtige, og de foretagne
konklusioner forkerte. Det må derfor også konstateres at syns- og skønsrapport BS99-1273/207’s
værdi, som et uvildigt og retvisende dokument er yderst begrænset.

Translation:

In the rapport by the engineer are referred to DS/EN60268-3, but it must be concluded, that a majority of the conducted measurements are based on a false assumption of this and other standards. This has lead to the measuring must be named wrongly, and the conducted conclusions wrong. It must therefore also concludes that the reports value, as an independent and true document is highly limited.

Now for the specific test Kurt von Kubik refers to, it stated that our B-110 would become to hot (raising more than 30K from ambient temperature = go from 20 degrees celcius to 50 degrees celcius). And that after first 1 hours preconditioning at full power, then after that a further 4 hours testing he would have to reduce the output power, and after 5 hour would have an output around 12 w in 4 ohm.

All other products tested, delivered the promised output over these 5 hours.

However –as mentioned before- the engineer had misunderstood the standard. There is currently 2 standards for amplifier rating: CEA490A (which we use), and IEC60238 (which the engineer used and misunderstood). These two standards claims for Power Output (NOT PMPO or any other fancy poweroutput rating) 5 minutes for CEA490A and 1 minute for IEC60238.

Now the test showed that our B-350 delivered full output in 4 and 8 ohms for 1+4 = 5 hours. And that our B-110 delivers full output in 8 ohms for 5 hours and 4 ohm for more than 1 hour. Both are significantly over what is needed according to any standard, and are with out any doubt also much more than many of our competitors. (most have a termal protection which will shut down if this test was conducted at them for 5 hour, should they then be rated for Zero watts ?).

However as mentioned Dansk Standard concluded that this was a falsely conducted test, as the temperature test is actually going to be conducted with the sinus attenuated 10db. (minus 10db on the input, is 1/10 on the output). AND because it was not a poweroutputtest, but a test on powersupply variations ! This means that the temperature test was completely wrong.

Dansk Standard specifically said:

I DS/EN 60268-3 paragraf 14.2.2.1, der omhandler strømforsyningsvariationer og fejlagtigt er
anvendt til udgangseffektmåling, er det specificeret at måleopstillingen skal være i ”standard
measuring conditions” som er specificeret i 3.1.3. Her fremgår det at indgangsniveauet skal
reduceres med -10 dB. Det betyder at temperaturmålingen der foretages, ikke foretages ved den af
fabrikanten opgivne effekt, men ved den effekt apparatet giver fra sig med et indgangsniveau der er
reduceret med 10 dB.

Translation:

In DS/EN 60238-3 paragraph14.2.2.1 which is about powersupplyvariations and wrongly are used for poweroutput measurements, it is specified that the measurementsetup should be in “Standard Measuring Conditions” which is specified in 3.1.3. Here it is stated that the input level should be reduced by -10db. This means that the temperaturemeasurement there is conducted, should not be conducted at the poweroutput rated by the manufacturer, but the poweroutput the product gives with an input level reduced by 10db.

So the conclusion is actually that our products performed exceptional when for 5 hours being forced to play an sinus at a level 10 times over the level required by the standard.


Further to my knowledge we do not have any customers, who sits and listen to a sinus for 5 hours, but in most cases they can do so, without problems…

Now I can only guess at the motives for these posting by Kurt von Kubik, and to a lesser degree from Hurtig.

Hurtig is aware that Kurt von Kubik did not win any cases and that the measurements was wrong, so all honesty would be to correct that in public.

For Kurt von Kubiks motives, they seems rather dubious, especially since he knows the measurements was wrong, and he is aware of the official statements from Dansk Standard.

However, if he tries to say that this test has any validity despite it being a falsely used test (refer to the official statement of Dansk Standard), there is only one thing to say, he claimed this for his amp:

Compared to my own amp, which is a "real class A" amplifier producing 100 Watts pr. ch. into 6 Ohms, weighing 100 Kg, running @ 55 dgr. C, QUOTE]

IF his amp was tested in the same way as ours, the temperature would raise above 50degrees celcius (55 according to his own writing), and therefore, the ONLY way his amp could pass this (wrongly conducted) test, would be to shut it down, resulting in the poweroutput being: ZERO !!! :-D

Best regards


Thomas Sillesen
Head Honcho Densen
 
Søren is amusing though, i love his various "IL Pape" avatars, also the one at the hifi4all forum.

The thermal layout of the B-110 output stage resembles that of the SAC 50-60 power amp.
The SAC "Igel" also sports a single pair output stage, 130W TOP3 devices, also mounted to the bottom of the amp case that serves as a heatsink.

The anodised aluminum mounting/heatspreader plate for the A1216/C2922 pairs of the B-110, custom order job, looks like a thermally well thought out solution.
Maybe Mr Sillesen can enlighten us how the heatspreader is connected to the bottom of the B-110 amp case.

SAC amp board=>
 

Attachments

  • Igel 50.jpg
    Igel 50.jpg
    21.3 KB · Views: 972