Amplifier Modules and PCBs For Sale

[Greg32]

Hi thanks for all the build info, quick question re mosfets your bill of materials says to only use IPP320N20N3 fets, your ebay add says irfb4227 are ok. are the 4227s ok to use as i already have stock of these. my rail voltage will be +-62v regulated. Thank you Greg

 

[Greg32]

disregard last post found alternative build info for irfb4227. greg

 

[Deleted member 148505]

Yup BOM for IRFB4227 build is in the 2nd page of the spreadsheet. Good for rails up to +/- 75V DC.

Disadvantage is that the IRS2092 will run hotter (~5 to 7 deg C), you need to put DIP16 heatsink. Heatsink will also run hotter because of higher current in mosfet driver.

 

[Greg32]

what are the main benefits of the optimum build? ive just looked up my local supplier and the IPP320N20N3 fets are quite cheap. thanks Greg

 

[Deleted member 148505]

For the benefits, you can run it on higher self osc frequency - 450khz to 550khz. Carrier residual is lower due to higher operating freq. Also you can set lower deadtime setting for better THD (but not really noticeable on sound). IRS2092 runs cooler and main heatsink is also cooler because of lower gate drive current. Around 2W to 3W will be saved in idle power consumption. Very good for active speaker systems, where the amp is inside the speaker box.

Because music has low crest factor, same average power can be achieved by IPP320N20N3 without heating up excessively.

However if you plan to use the amp in 2 ohms load or bridged mode 4 ohms and use it in pro audio (highly compressed music), IRFB4227 is better.

Don't substitute the values of critical parts in IRFB4227 build, that build has been tested, hundreds of my JLA-250D modules are running with IRFB4227 without problems.

 

[Greg32]

I will be using the T130-2 toroids how many turns for 22uh thanks Greg

 

[Deleted member 148505]

I will be using the T130-2 toroids how many turns for 22uh thanks Greg

45 turns, 195cm long. Use 18 AWG.

 

[Deleted member 148505]

Dip it on insulating varnish, and again after it finished curing to prevent the wires from vibrating.

You can also use 1D31A-220MC available on Mouser

 

[Greg32]

Thanks, Greg

 

[Greg32]

Hi Lester I see in one of your previous designs you use a litz wire on the inductor is there any benefit or is std 18g enameled wire ok at audio frequencies. thanks again Greg

 

[dflorin]

Hi. JLE-1500D not for sale on eBay. +/- 80V supply voltage dc and speaker impedance 4ohm
Final transistors can move laterally using wires? I want four final transistors (bridge - for safety)

 

[Deleted member 148505]

Hi Lester I see in one of your previous designs you use a litz wire on the inductor is there any benefit or is std 18g enameled wire ok at audio frequencies. thanks again Greg

If you use Litz wire for class-d, the inductor will run cooler, it has lower heat losses at frequencies 10khz to 1ghz. Enameled "Solid" wire is also good if you don't have litz wire because high current signals that are passing through the inductor are audio frequencies. Principal loss at low frequency signals is due to winding resistance (DCR)

Hi. JLE-1500D not for sale on eBay. +/- 80V supply voltage dc and speaker impedance 4ohm
Final transistors can move laterally using wires? I want four final transistors (bridge - for safety)

Still acquiring lightweight aluminum heatspreader. Readily available angle bar here will exceed 500g shipping weight. If the package exceeds 500g the shipping price will double. (pricing for +500g to 1kg are almost the same)

JLE-1500D can drive 2 ohms load with +/-80V DC supply, you don't need to add mosfets. Because music has a high crest factor, worst continuous output is only 1/4 of the headroom, the amplifier module can handle it. Just attach the heatspreader to a large heatsink and provide additional cooling if you want to run it whole day at near clipping conditions.

If you add or extend mosfets through wires, the switching time will be affected due to stray inductance. If you are lucky, only the THD will be affected. Common scenario, the mosfets will heat up and each mosfet will heat up at different temperature. Also the highest power that can be attained will be lowered because one mosfet might heat up badly and will make the amp fail. Worst, the overcurrent protection will be falsely triggered at low power output rendering the amp useless.

 

[Deleted member 148505]

THD Spectrum measurements of JLA-250D in a typical (non laboratory) setup. No fancy/novelty parts.

Linear power supply setup
KBPC3510 rectifier,
15000 uf caps
+/-53V DC Rails.
12VAC supply for speaker protect circuit.

JLA-250D V1.2 setup
Gain: 39
Deadtime: DT3

JLE QA400 / Soundcard Interface setup
+/-12V DC rails
RCA Cables with adaptor

8R 1W

8R 5W

8R 10W

8R 50W

8R 100W

 

[Greg32]

Hi Leister can you advise on R21 on the jle800 i cant find 75r locally but 68r or 82r are available (phillips pr02) would either of these work?

 

[Deleted member 148505]

Hi Leister can you advise on R21 on the jle800 i cant find 75r locally but 68r or 82r are available (phillips pr02) would either of these work?

Yup it will work.

Regards,
Lester

 

[Greg32]

Cheers thanks

 

[Deleted member 148505]

Coming soon, PCBs for...

Mauro Penasa's Myref (without the power supply)
Designed to be used with SMPS with +/- 15V gnd referenced aux and +15V or +12V floating aux supplies.
PCB Size: 2.1 x 5.0 in

Adapter for bridging amplifiers.
That1206 balanced receiver with dual opamp buffer for inverting and non-inverting output,
used for bridging amplifiers to get more output power.
PCB Size: 2.5 x 2.7 in

 

[Deleted member 148505]

I've read some reviews that pre-filter feedback type class-d's do sound more tube-like than post filter feedback types. I'll try to compare the two by creating post feedback implementation of IRS2092, it is based on LiteAmp by Chocoholic.

I've changed the input and power supplies used, mine only has single ended input buffer. External VDR and +/-15V supplies. open collector nfatal signal, mute by pulling to ground by external trigger.

Some 3d rendering...

 

[Greg32]

Hi Lester, Its been a slow build due to waiting for parts to arrive, anyway i had no problems with either board both sound really nice and detailed and not bright or harsh, credit to you for a quality design. my rails are +-60v and im getting about 130w at 8r with a 100hz sine wave at just below clipping. Greg

 

[AlexScan]

I've read some reviews that pre-filter feedback type class-d's do sound more tube-like than post filter feedback types. I'll try to compare the two by creating post feedback implementation of IRS2092, it is based on LiteAmp by Chocoholic.

I've changed the input and power supplies used, mine only has single ended input buffer. External VDR and +/-15V supplies. open collector nfatal signal, mute by pulling to ground by external trigger.

Some 3d rendering...

Good to see that Jan !

I think your "quality made" module associate to the pre-filter schema of chocoholic can make a magic result .

I read the post of the SystemD LiteAmp, impress by the result without "quality components" @ 10W/4R: -91.1 dB / 0.0028% THD

Comparing to your measurement of the JLA250/800 it is the same (normally at 8R => less THD than 4R), but it's just measurement, can't replace serious listening !