Has anyone tried this "350 watt" amp

Status
Not open for further replies.
Jump to Latest
#2
I don't know the amplifier nor what components have been used.

The 4pr output stage should be robust enough to do 200W into 8r0 and into 8ohm
If 1943/5200 are used I would not use a +-70Vdc PSU.

I think the 4pair output stage could survive driving a 4r0 load but I cannot have confidence that it can drive 4ohm reliably. If MJL4301/4281 are used I would have much more confidence in ONsemi reliability and this assembly.
 
#3
Most of these cheap kits on ebay are way overrated. I'd say this is a 100 watt into 8 ohms and 200 watts into 4 ohms amplifier on 50V rails.

I doubt their even tested into claimed loads and that claimed wattages are just based on number of outputs and the supply voltage.

If an amp cannot double the wattage into 4 ohms, its not really 4 ohm stable.
 
#4
Tekko said:
If an amp cannot double the wattage into 4 ohms, its not really 4 ohm stable.
Click to expand...
A really good audio amplifier, even a ClassA version cannot double it's power output when the load resistance is halved.
Cordell's "rule" is 180% power into halved load resistance.
I used to aim lower than this, but have revised my modeling methods to achieve Cordell's rule.

driving a 4r0 load is no guarantee of stability.
doubling power in halved load resistance is no guarantee of stability.

If this amp actually tested at 200W into 8r0 and 350W into 4r0 then that gets very close at 175%, but that does not mean it can drive a 4ohm speaker.
Resistive test loads prove very little, other than current delivery.
 
#5
Because even the best non-regulated power supply and output transistors do not have a 0 ohms resistance at clipping, and also because of the emitter resistances, you will never see exactly double the power in 4 ohms as in 8.

Also, despite that variation most of the difference between the 4-ohms and double-the-8-ohms measurements are due to a sagging power supply.
 
#6
AndrewT said:
If MJL4301/4281 are used I would have much more confidence in ONsemi reliability and this assembly.
Click to expand...
Wow, MJL4302 PNP are over $6 each today, where as MJL4281 are only $4.50. What makes these better than MJL21193/4G which are $4.13 and appear to have more SafeOperating Area at 80V. I realize 4302/4281 are TO264 and 21193 are TO247, is the TO247 to hard to fit in the kit under discussion?
(datasheet from datasheetcatalog says MJL4302 is complement to MJL4281).
 
#7
Tekko said:
Most of these cheap kits on ebay are way overrated. I'd say this is a 100 watt into 8 ohms and 200 watts into 4 ohms amplifier on 50V rails.

I doubt their even tested into claimed loads and that claimed wattages are just based on number of outputs and the supply voltage.

If an amp cannot double the wattage into 4 ohms, its not really 4 ohm stable.
Click to expand...

4 pairs = 100w ?? Maybe if fake semi's are used. I do agree the components may be cheap , but the boards are worth it. BTW, 2 pair njw0281/0302's can do 120w ALL DAY 8R- 200W/4R. I saw the design for this one , basic fully complimentary front end- standard EF2 ... should sound "accurate" (good bass). I won't go beyond 2 pair now without a triple , 4 pair on a EF2 sucks for 4R.

ALL ebay chinese class AB amps = VERY basic designs , no "frills" (cap multipliers , TMC , accurate CCS's) many lack base-stoppers , rail decoupling, some even oscillate !! 😱 .... most are quick ripoffs of OEM's or DIY designs ... really.

OS

OS
 
#8
indianajo said:
Wow, MJL4302 PNP are over $6 each today, where as MJL4281 are only $4.50. What makes these better than MJL21193/4G which are $4.13 and appear to have more SafeOperating Area at 80V. I realize 4302/4281 are TO264 and 21193 are TO247, is the TO247 to hard to fit in the kit under discussion?
(datasheet from datasheetcatalog says MJL4302 is complement to MJL4281).
Click to expand...

Get NJW0281/0302 $2.10 each ($1.90/25 mouser) ... 90% of the MJL SOA. I have not blown one yet in real world use. 200+ repairs with this pair from ON semi - 0% failure. These are so cheap , I use them as the drivers in triple OP's (cheaper than MJE15032/33(to-220).

4 pairs of these excellent devices run paralleled 8R woofers in my local church (over 1 year now) 250-300W easily.

OS
 
#9
I have some experience assembling and using local kits available here for several years now, based on five pairs of MJL21193/4. Intended for DJ work, they have good basic reliability but can still manage a genuine 350W/8 ohms, 500W/4ohms with great fidelity.
There is no apparent reason that an amp using four pairs may not produce an assumed rating of 350W/4ohms. provided similarly rated components are used, plenty of copper on the PCB and a substantial 0.4 deg.C/watt heatsink is used for optimal bias in class AB.

Rail voltages typically must be reduced to suit the lower impedance for safety and obviously, this will reduce the 8 ohm capabilities unless you can tolerate poor audio from current limiting action at 4 ohm loads. This will be just one of the compromises likely with $38/pr amplifiers. Compared to the likely total amp cost, it's peanuts so why skimp on the heart of the investment?
 
Last edited:
#11
ostripper said:
Get NJW0281/0302 $2.10 each ($1.90/25 mouser) ... 90% of the MJL SOA. These are so cheap ,


OS
Click to expand...
Cool. I have just installed 5 each NJW21193 NPN On Semi at $1.93 Newark, but searched and searched their tables for TO3-P 15-16A PNP transistors and it didn't come up with NJW0302. But they have them @ $2.13. This will save me enough over MJ21195 ($3.73) I won't have to wait until June to buy them. Due to not having chassis diagram, just PCB diagrams of PV1.3K, I mis-identified blown PNP-s as blown NPN's and bought the wrong transistors. Red wire is plus 95V (NPN), yellow wire is -95V(PNP), the colors should have been obvious (not).
 
Last edited:
#12
In the church , I use a 600VA 50-0-50Vdc setup , the 4 pair NJW's have awesome SOA at 50V and below with a large heatsink , almost 21193/4 performance (6.5mm dies) , easily does 300W/4R ..... no issues , will last 20 years. 🙂 Only difference , 8 devices = only $15USD !!! Can't beat that. 🙂

With these kits , I would spend the 15$ , also replace the caps with well derated (100V- 105c) Nichicon/panasonics , maybe even use the fairchild small signal devices (ksc1845/a992 - 3503/1381). as i said , just the boards are worth 38 bucks.
Also , I would sim the heck out of the circuit , followed by the CRO/squarewave tests into load - tweak compensation / dethump the CCS's ... among other things "do the magic". Once you have boards , all the rest is "fun/play" 😀

OS
 
#13
indianajo said:
Cool. I have just installed 5 each NJW21193 NPN On Semi at $1.93 Newark, but searched and searched their tables for TO3-P 15-16A PNP transistors and it didn't come up with NJW0302. But they have them @ $2.13. This will save me enough over MJ21195 ($3.73) I won't have to wait until June to buy them. Due to not having chassis diagram, just PCB diagrams of PV1.3K, I mis-identified blown PNP-s as blown NPN's and bought the wrong transistors. Red wire is plus 95V (NPN), yellow wire is -95V(PNP), the colors should have been obvious (not).
Click to expand...

NJW0281G ON Semiconductor Bipolar Power

They still have 408 of them , just bought 25. 😀

NJW0302G ON Semiconductor Bipolar Power

Still have 547 of these , just bought 25 . (wow $1.73 each)

PS - in lots of 25, Hfe is within 1-3 in the group - very easy to match , almost "prematched" VERY consistant manufacturing. (newark sucks - I switched to mouser 2 years ago - they ship to me in 48 hours / texas to tenn. = $7.10 UPS)
OS
 
Last edited:
#14
ostripper said:
NJW0281G ON Semiconductor Bipolar Power
They still have 408 of them , just bought 25. 😀
NJW0302G ON Semiconductor Bipolar Power
Still have 547 of these , just bought 25 . (wow $1.73 each)
OS
Click to expand...
From here mouser is $9.50 UPS or $9 USPS, Newark is $6 UPS. It matters in lots of 5 and on my income these days. This is the first amp I have worked on that has more than 2 each TO3/channel. I wanted a $99 CL PV4 but Duck river Tennessee was too far to drive. I found this PV1.3K @ $55 on the Louisville bus line, a bargain for the power transformer & heat sinks even if all the semi's were blown. Only 1/3 of them were. Probably horrible at 1W where I listen in the L.R., but just the thing for two quad 15" black widow bass walls of sound, if I needed such a thing.
 
#15
I have built an amplifier using the L20 boards. Details can be found in the Solid State gallery. The transformers I used probably precluded 350 wpc. I would guesstimate 150 wpc. The sound is superior to a Yamaha M-70 that it replaced.
 
#16
Thank you all for L20 attention.

The EF3 is L20. Working voltage is + 65 V 200 W8R.


It can be compatible with the ONSEMI NJW0302G transistor. And there's not much difference.

L20 is very popular in China, but please note. Good things there will always be people who will go to copy.

Many of the PCB is red compared inferior components.

In China there are many such businesses do. I feel sad.


Here, you can see some of the specific material. L20

At present, L6 L20. Is China's most popular amplifiers, it is not for power.


But has the most powerful control. Driving the hardest speakers. And not used to the biggest volume.



L20 ÖÕ¼«°æ¡£_·Å´óÆ÷°æ_LjmAudioÒôÏì¼¼ÊõÂÛ̳


An externally hosted image should be here but it was not working when we last tested it.
 
Status
Not open for further replies.
Top Bottom