Once upon a time, like 4 decades ago, 2 x 1000W from a 2 Units of height amp seemed impossible and unaffordable like a new world wonder. Class D has changed the situation dramatically - today there are many Chinese companies where you can get this power for ~USD 300 plus minus a few bucks for shipping and duty... Please hold in mind that I did not say that the sound quality of the chinese stuff is good or superior to old class AB designs - just that the watts per dollar plus the watts per cubic inch have increased dramatically.
On the side of the "miracles" today, there are new things the idea / imagination of which would have been ridiculed 40 years ago like the concept of air travel would have been ridiculed within the 18th century.
One of these things is the KW18. I personally know a guy who is not an audio charlatan, but more or less the opposite. He is into design / build and employment of a 5 way horn system for installations and larger outdoor crowds. The system is called the Lambda Labs QX-3. I ve been at a demo show and I have the feeling that I ll never be able to afford such a system, not just because of the 6 figure system cost, but mainly because of the 7 figure cost of a suitable venue where to home it (the average livingroom or home discotheque won't suffice ;-).
Anyway, he has developed a 2 rack-units amp too ("KW18") , that he's claiming to put out 4 x 4500W into 4 Ohms using 4 x UCD2k full bridges at ~ 190VDC.
The PSU rail must be super-stable with almost no sag even at an average load of ~100Amps, with the peak load current at sine peak approaching almost 200Amps for very short time periods.
Is it possible to put 4x UCD 2k plus adequate forced cooling for overloading each of them to 4.5kW PLUS this kind of mega-PSU into a 2 U case?
What is your opinion?
(His website lamda-labs.com is outdated, and he is unfortunately showing current updates on a social media site which I do not use / will never use nor do I mention it because their strange policy on free speech)
On the side of the "miracles" today, there are new things the idea / imagination of which would have been ridiculed 40 years ago like the concept of air travel would have been ridiculed within the 18th century.
One of these things is the KW18. I personally know a guy who is not an audio charlatan, but more or less the opposite. He is into design / build and employment of a 5 way horn system for installations and larger outdoor crowds. The system is called the Lambda Labs QX-3. I ve been at a demo show and I have the feeling that I ll never be able to afford such a system, not just because of the 6 figure system cost, but mainly because of the 7 figure cost of a suitable venue where to home it (the average livingroom or home discotheque won't suffice ;-).
Anyway, he has developed a 2 rack-units amp too ("KW18") , that he's claiming to put out 4 x 4500W into 4 Ohms using 4 x UCD2k full bridges at ~ 190VDC.
The PSU rail must be super-stable with almost no sag even at an average load of ~100Amps, with the peak load current at sine peak approaching almost 200Amps for very short time periods.
Is it possible to put 4x UCD 2k plus adequate forced cooling for overloading each of them to 4.5kW PLUS this kind of mega-PSU into a 2 U case?
What is your opinion?
(His website lamda-labs.com is outdated, and he is unfortunately showing current updates on a social media site which I do not use / will never use nor do I mention it because their strange policy on free speech)
Last edited by a moderator:
As long as you can find power-FETs that can stand the voltage (you will need many in parallel) and you match a class D amplifier with a correspondingly powerful power-supply, you can get KW output power.
But what useful results does that bring? The sound will probably be inferior to an amplifier with less power because you may have problems with the switch-over between the low- and high-side switches if you cannot find a monolithic driver that can handle the voltage and the gate-capacitances to drive.
KW power is quite useless for practical purposes. You may have serious issues with heat inside the amplifier.
A victor for mankind?
But what useful results does that bring? The sound will probably be inferior to an amplifier with less power because you may have problems with the switch-over between the low- and high-side switches if you cannot find a monolithic driver that can handle the voltage and the gate-capacitances to drive.
KW power is quite useless for practical purposes. You may have serious issues with heat inside the amplifier.
A victor for mankind?
... For a few miliseconds (or maybe seconds), until all the protection circuitry kicks in and limits the output so the thing doesn't blow itself up... 
Psoft K20 only 2 channels - 2 halfbridges running on about 2 x 225Vdc. They use probably 500V or even600V Fets. Completely different story.
With the Kw 18 the issue is the narrow limitsfor the 190V bus since the UCD fullbridge uses 200V components. To ensure the power, sag must not go below 190Vdc,but all DCbus ripple, negative load step regulator overshoot plus switching overshoot must NEVER exceed a total of about 196V at the drain of the Fets - including all unit by unit tolerances (! ) to keep at least a 2% voltage derating, what is anyway too less for reliable operation.
With the Kw 18 the issue is the narrow limitsfor the 190V bus since the UCD fullbridge uses 200V components. To ensure the power, sag must not go below 190Vdc,but all DCbus ripple, negative load step regulator overshoot plus switching overshoot must NEVER exceed a total of about 196V at the drain of the Fets - including all unit by unit tolerances (! ) to keep at least a 2% voltage derating, what is anyway too less for reliable operation.
Last edited by a moderator:
I wonder how 4,5k Watts at 4 Ohms should come out of two "bridged" Modules, since the UCD2K is a full bridge already and is not optimised for 2 Ohms.... This in no way can mean "4500 Watt @ 4 Ohm at full sine for minutes" but rather the typical 18 dB Crest for a bunch of ms.... which totaly is possible and done by many others around.. Besides - 4 x 4500 Watt sine would require more power than the typical 16Amp sockets could deliever..
This of course could deliever more power, but still much is unclear.
And how did he bridge the modules ?
Which PSU is he using?
Which crest factor is the wattage based on, for how long, what frequency, etc...
Claiming to have this much power is easy , there are amps with specs like this in the datasheet (but of course its all peak, high crest, etc...) - having real 4 x 4,5kW @ 4 Ohm long term sine seems *a little* unlikely... Doing it with the Hypex Amp ... see questions above.
If you need this much power, why not simply buy a PKN 3 Phase ....
And how did he bridge the modules ?
Which PSU is he using?
Which crest factor is the wattage based on, for how long, what frequency, etc...
Claiming to have this much power is easy , there are amps with specs like this in the datasheet (but of course its all peak, high crest, etc...) - having real 4 x 4,5kW @ 4 Ohm long term sine seems *a little* unlikely... Doing it with the Hypex Amp ... see questions above.
If you need this much power, why not simply buy a PKN 3 Phase ....
Last edited:
This is all possible. Take a regulated PSU, have 2x95Vdc regulated to feed the UcD2K and you will have 4,5kW in 4 ohm.
I know the customers (of course...), you will be impressed with what they have achieved. As far as I know it will be one of the most powerfull amplifiers what runs on one mains voltage.
Please contact them direct, obvious I can not give any more details.
For the customers who thinks we are not Pro Audio focused, this proves we are...
I know the customers (of course...
), you will be impressed with what they have achieved. As far as I know it will be one of the most powerfull amplifiers what runs on one mains voltage.Please contact them direct, obvious I can not give any more details.
For the customers who thinks we are not Pro Audio focused, this proves we are...
Last edited:
Thing is, nearly nobody doing large scale pro audio really cares about long term sine wave rating, because it really tells you nothing very useful.
I mean I care about the amp having a sufficiently large long term power handling, but I ALSO care about the short term (ms to hundred ms or so maybe) being very much larger then the continuous rating because that actually matters to avoiding clipping.
Rule of thumb is that long term average is only about 1/8th of the power of the corresponding short term peaks, so having 10db of difference between the two is actually quite reasonable and makes for a much lighter amplifier.
Give me 1kW long term with sufficient headroom to deliver 10kW peaks for a few tens of ms at a time and a few percent duty cycle and I will take it in preference to a 2.5kW amp that smacks into the supply rails at 2.6kW any day of the week.
Regards, Dan.
I mean I care about the amp having a sufficiently large long term power handling, but I ALSO care about the short term (ms to hundred ms or so maybe) being very much larger then the continuous rating because that actually matters to avoiding clipping.
Rule of thumb is that long term average is only about 1/8th of the power of the corresponding short term peaks, so having 10db of difference between the two is actually quite reasonable and makes for a much lighter amplifier.
Give me 1kW long term with sufficient headroom to deliver 10kW peaks for a few tens of ms at a time and a few percent duty cycle and I will take it in preference to a 2.5kW amp that smacks into the supply rails at 2.6kW any day of the week.
Regards, Dan.
Yes, long term sine wave. The limiting factor is how much heatsink and forced airflow you have applied. With good cooling 1/3 to 1/2 power continues is possible, in practice 1/4 to 1/3 continue power is enough. Nobody listen for a long time to sine waves...
. Bridge is not possible, the UcD2K is already a bridged amplifier.I am well aware, that sine wave is not a real world figure But it helps to get a better picture of the heavy bass duty performance an amp is capable of. If detailed measurements with crest factors and duration were posted, I wouldn't ask
One good example where its obvious that peak is not everything and that the lack of long term power is very noticable with heavy bass music is the powersoft X4/8. There is enough talk about that over at speakerplans, just wanted to point out why I am asking...
@Jan-Peter
Thank you. This clears up the initial question of bridging two modules in Post 1 which is not possible (UCD2K already being bridged), as I remembered.
This sounds good - and brings back an old idea - I see you wrote me a PN... let´s talk
But it helps to get a better picture of the heavy bass duty performance an amp is capable of. If detailed measurements with crest factors and duration were posted, I wouldn't ask One good example where its obvious that peak is not everything and that the lack of long term power is very noticable with heavy bass music is the powersoft X4/8. There is enough talk about that over at speakerplans, just wanted to point out why I am asking...
@Jan-Peter
Thank you. This clears up the initial question of bridging two modules in Post 1 which is not possible (UCD2K already being bridged), as I remembered.
This sounds good - and brings back an old idea - I see you wrote me a PN... let´s talk
Jan-Peter, of course 190Vdc can give 4.5kW. BUT:How high is the overshoot during switching,? How can a bridge of 200V components suffice for that rail voltage? Even worse, if the 190Vdc rail(or 2 x95V) is not controlled tight enough and may at times of no load reach a few Volts higher? Why is UCD2k called UCD2k and not UCD4k?
Last edited by a moderator:
With a tight controlled PSU you can achieve this. The first peak will come from decharging the electrolytic capacitors, obvious you will need many.
We have rated the module to have this power in 4 ohm AND 2 ohm load in combination with the SMPS3K. With a regulated PSU you can have much more. However at that time we didn't had this. With a tight mains and with one SMPS3K and one UcD2K you can get short time peaks of 3,2kW in 4 ohm. At the end of the day it's just a name, we don't like to overrate our products. We prefer to be realistic
.
Last edited:
Some work is required for getting the sense of pro-audio: rigging various equipment at random locations for some years, doing live mixing of various types of art for some years, doing component-level repair of random equipment (pro-audio and not pro-audio) for some years, collaborating randomly in art-related events for some years, and all related environmental study.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.