Nobody is accusing you of anything , and allot of things can go wrong assembling a board , you should know ! so there is no case here .
I assume you're an experienced electronic guy , so if you want to make a honest opinion , just order some LJM's from an "Authorised source" put them together and then make your judgement .
Cheers ,
Rens
I too would like to understand the failure of the ljm amps in question. It may or may not be related to sub-standard parts. Maybe a direct short. Maybe driven to clipping and beyond. I appreciate your willingness to do this. My point was not to accuse anyone of anything, but in the interest of full-disclosure the next time you should probably mention that you sell a competing product.
I have purchased 2 SMPS from Connexelectronic and I'm very pleased with the quality and performance of the units. I too, believe there's alway room for improvement in the level and quality of amps I build. I will very much consider looking at what you have to offer for a future project.
Rick
I understand that there must be a price threshold, below that, the quality of the parts doesn't matter at all or is seen as "fine for the price" or even excellent. Then once we cross that threshold the expectations for parts quality rise exponentially. Or simply my expectations and quality standard are far too high? I'm just curious what's that quality threshold. If is not that low, I guess I can substantially increase my revenue by just decreasing the quality level for my boards while keeping the price, or keep the quality and rise the price. Eventually the price/quality would be the same as I discover now that the market ask for.
All the so called allegations are based on objective observations, and can be verified following this thread and pictures from the post mentioned above, as well as by those who have purchased similar boards as the ones I got my hands on. Or there is a reason for them to keep quiet.
All the so called allegations are based on objective observations, and can be verified following this thread and pictures from the post mentioned above, as well as by those who have purchased similar boards as the ones I got my hands on. Or there is a reason for them to keep quiet.
Same here, pair of L20D driving Econowave passive crossover speakers - driven by a Connex PSU.
I did apply the recommended tweaks mentioned at the start of this thread... applying the missing heat sink paste, low ESR filter caps. I have never run them near their full capability, I just like having the headroom there.
We can argue about the diminishing returns of the quality of the parts used in any amp. But, you also stated in post #557 that the design and layout of the L-series amps pcb is not just sub-par, but terrible for class D. Is this just a difference of approaches of design implementation between two 'engineers' (ljm and Cristi), or do the items you mention in your post affect the overall, noticeable, audible and sonic characteristics of the amp? Operative word being noticeable. Apparently, the sound must be 'good enough' based on testimonials in this forum.
Extra capacitors only for bass L15D-PRO?
A quick question to those who know much more than I do
I am now building my 3 channel L15D-Pro based amp with the Connecs smps500R at +/- 50V.
My idea is to "insulate" the high and mid boards from the bass board by putting 2 capacitors in the bass board power-supply lines. Hope this will flatten power fluctuations in the higher frequencies and improve sound stability (image etc.)
Heavy power off pops are not desired.
I can choose between 4700uF-80V or 10.000uf-63V or nothing at all of course.
Thank you in advance
Peter
A quick question to those who know much more than I do
I am now building my 3 channel L15D-Pro based amp with the Connecs smps500R at +/- 50V.
My idea is to "insulate" the high and mid boards from the bass board by putting 2 capacitors in the bass board power-supply lines. Hope this will flatten power fluctuations in the higher frequencies and improve sound stability (image etc.)
Heavy power off pops are not desired.
I can choose between 4700uF-80V or 10.000uf-63V or nothing at all of course.
Thank you in advance
Peter
Last edited:
A quick question to those who know much more than I do
I am now building my 3 channel L15D-Pro based amp with the Connecs smps500R at +/- 50V.
My idea is to "insulate" the high and mid boards from the bass board by putting 2 capacitors in the bass board power-supply lines. Hope this will flatten power fluctuations in the higher frequencies and improve sound stability (image etc.)
Heavy power off pops are not desired.
I can choose between 4700uF-80V or 10.000uf-63V or nothing at all of course.
Thank you in advance
Peter
I can't remember exactly, but I think it was Cristi who suggested adding around that amount of capacitance on the output of his SMPS was always a good thing. I'm using 2 L15D's with one of his 300R's and 6800uF/80V on each power rail. I can't say I did an A/B of with and without, I just built it that way the first time. (and no power pop offs)
Another happy user of Ljm's boards.
Have 2 L20d on a Connexelectronic SMPS driving two Markaudio speakers.
I'm actually getting better results (with my specific setup) than previously with a TK2050.
The boards (stock ones from one of the two recommended sellers) are well built, components are good for the price range, I've added an extra speaker protection driven from a different PSU.
The only serious tests I've seen are the distortion tests made by Trevmar here, the alegations made in this thread are for me just hearsay without any value until things are proven.
I'm sure Cristi can show us more detailled infos and tests (I know he is a skilled engineer, offering some great products) to help the debate here.
Until then, I continue very happy with ljm's L20d.
Have 2 L20d on a Connexelectronic SMPS driving two Markaudio speakers.
I'm actually getting better results (with my specific setup) than previously with a TK2050.
The boards (stock ones from one of the two recommended sellers) are well built, components are good for the price range, I've added an extra speaker protection driven from a different PSU.
The only serious tests I've seen are the distortion tests made by Trevmar here, the alegations made in this thread are for me just hearsay without any value until things are proven.
I'm sure Cristi can show us more detailled infos and tests (I know he is a skilled engineer, offering some great products) to help the debate here.
Until then, I continue very happy with ljm's L20d.
Yes, you're voltage limited unless you're operating BTL (which I'm not sure whether the L15D is capable of, and certainly not into 4 Ohms). The math is pretty simple.
Pout = Vrms^2/R = Vpeak^2/2R
Vrms = Vpeak/sqrt(2)
At clipping (perhaps 10% THD), Vpeak might be ~95% of the supply, so let's assume Vpeak = 40V
Pout = 40^2/(2 x 4) = 200W
voltage is one end of the question. The other is current.
P=R*I²
300Wrms in 4 ohms means current is 8.66Arms, ie 12A peak
I wonder why L15D is advertised for up to 350W in 4ohms as the mosfet IRFI4019H absolute max rating is 8.7Amp (down to 6A when temperature goes up).
@darpmalone : look a the definition of RMS value for a sinusoidal. You'll get it
P=R*I²
300Wrms in 4 ohms means current is 8.66Arms, ie 12A peak
I wonder why L15D is advertised for up to 350W in 4ohms as the mosfet IRFI4019H absolute max rating is 8.7Amp (down to 6A when temperature goes up).
@darpmalone : look a the definition of RMS value for a sinusoidal. You'll get it
If you were delivering DC power then you would be closer to correct. With ±42V supplies, you can deliver a 40V peak AC sinewave, so for 1/2 cycle, the power comes from the +42V supply, and for the other 1/2 cycle, the power comes from the -42V supply. Each supply has to be able to deliver 420W peak to deliver 400W peak/200Wrms to the load, but the RMS or average power from the supply is much less. The fact that each supply rail delivers power for only 1/2 the cycle is a reason that charge pumping can be such a problem. Operating dual outputs in BTL mode or out of phase for stereo channels keeps the loading on both supplies more constant for the full phase. If course, I am basing my comment on a single frequency sinewave, but it's generally applicable for any AC waveform except that the peak to average ratio becomes even higher than sqrt(2). This is why it's advantageous to buy a power supply that has been designed for audio applications. The thermal capability of the supply needs to be based on the RMS power delivered, but the peak capability should be much higher. Otherwise, you're overdesigning a supply to handle the peak, adding cost, size, and weight. Cristi appears to understand this, and his Connexelectronic supplies do this well. It's not such an issue with an unregulated linear (transformer) supply, since the solution is generally just to add more capacitance to handle the peaks without excessive undervoltage.Thanks for the reply cwier60. Just trying to get this str8.
I was thinking (hoping) that because I'm using dual voltage (+/-42v) I would see 200W/rail. How does the second rail come into play, power-wise then?
I've always been confused about this
BTW, if you had a single 42V supply, then your peak-to-peak, as opposed to peak, AC voltage would be 40V, so your output power would be 1/4.
I have no idea what the IRFI4019H devices in the L15D can safely handle, but the 8.7A rating is for continuous DC operation. Any single device in the bridge sees much less than the rated output current. The power rating has more to do with the thermal design than the current rating, since for most MOSFETs utilized in switching applications, only the pulsed current rating means much, and that is still very dependent upon the application.
agreed. Pulse rating is 34A for this mosfet.
It is difficult to estimate the pulse current really used, at least to me.
Maybe I overlook this : In my mind, even for switched applications, RMS current needed to deliver power in a load is an average which can be considered as equivalent to the constant current the mosfet should be able to deliver safely. Does it make sense or is this a wrong assumption ?
Music has less energy than a sinusoidal with same peak. Maybe would be ok for music.
Perhaps I'll explain what I'm doing... I'm building dual mono amps using 2 L15D Pro boards. I've built two linear PS's consisting of two 60VCT/450VA Paradigm transformers & 20,000uF/rail. So I've got +/-42VDC rails. (LJM recommends +-35V to +-50V for this board)
I figured +-42V would be a great compromise between 4R and 8R speakers. In-fact I thought I was over-designing the power supply. If I'm voltage limited with +-42V @ 4R, how is LJM recommending +-35V to +-50V?
Thanks again
300Wrms in 4ohms can be reached if power supply is +-50Vdc. (and if power supply can deliver the current and if mosfets can sustain such a current)
interesting also is that the pro board, same as previous one with onboard speaker protection, has a larger heatsink. He indicates that this is to be able to reach 300W in 4 ohms. The previous board, with smaller heatsink, was supposed to reach 350W in 4 ohms. Suspicious ?
interesting also is that the pro board, same as previous one with onboard speaker protection, has a larger heatsink. He indicates that this is to be able to reach 300W in 4 ohms. The previous board, with smaller heatsink, was supposed to reach 350W in 4 ohms. Suspicious ?