Power amp under development

Jump to Latest
#3,482
Hello, some time ago i had problem with this amp module. It was burning Fets " without any reason". I decided to deal with thi problem. First problem was with unmatched Fets. They can differ very much so you set current for this stage, and 3 of 4 of them should not begin to work at all, but only one take about 120 mA. I matched Fets-they take current with max 30% dispropotion-not bad. But i think there's bug. Current was set at 10R resistor in place of fues. I decided to set it on lower level-to protect fets during tests. This time i should place there fuses and there always by amp is burning, but i decided to give additionaly another 10R resistor-which give us about 5R totally. Current was set to achieve 0,8 V before so now should be about 0,4V-there's problem, when i started amp with voltometer there was about 0,6V @5R. I decreased it to 0,4V and inserted onother 10R resistor. It gives us about 3,3R and measured voltage should be about 0,26V. Nothing like this, it was 0,46V. There is problem with this module. When i put there fuses (0R) current rises to very big value causing unstability. Why it happens? Rails are 2x71VDC without load, under load of muted amplifier it falls to 2x 69VDC. I use 4 pairs of IRFP250N, MJE 15033/34 for driver and predriver (there are different leg layout-they are reversed), MPSA for all BC's and 2SA1845 in differential stage. Temperature compensation works very well. After few seconds total current is falling about 5-10mA (predrivers, and drivers placed on main heatsink, with thermal transistor, insulated by pads). Main project is copy of quasi's design with changes given.
 
#3,485
DC detect schematic

You can but it is not ideal. Attached is a DC-detect cct you can use. It is a variation of one presented by a number of electronics magazines here in Australia.

Cheers[/QUOTE]

Hello Quasi
I'm looking for a good quality DC protection and I found your schematic on the forum (look like a very good design and simple!)I would like to use it in my newly built amp. I have DC problem when I turn off my amp, 6-7VDC goes out to the speaker terminal. When the power on the amp works well and only 20mV offset .
I do not understand something .
On the schematic there is a amp ground but on the PCB layout (silkscreen)I only see the 24V power supply ground ?
Or the amplifier ground and the 24V PS ground have to be connected?
Would you please let me know .
Without the proper connection probably not working how it should . Or degrade the sound of the amplifier or it can be worst I burn the amp on the first turn on .
Thank you 🙂
mathelaszlo2@yahoo.ca
Greeting
 
#3,488
Thank you very much Quasi

How do these DC detect effect the sound of the amplifier ?
Did someone tested with the DC detect circuit and with out the DC detect ?
Somebody advised to me to use a large capacitor on the speaker output .
I don't know which one would be better ?

Greetings
 
#3,489
Why dont you use CROWBAR for DC detect and electronic turn on delay / fast off in NMOS ? Its very safe and easy to do.

A lot of Pro Audio Amplifier have CROWBAR and use this way

Have built 3 pcs. NMOS 350/500 and its working very good, very very good sounding and stable working amplifier

Have add some features in latest NMOS design like Balanced Input with OP NE5534, Input Limiter and VI Limiter + CROWBAR DC detect

I use in Input Stage for BC - MPSA 92/42 and 2 SC2240
 
#3,490
quasi said:
Hi Qsa,

Sorry but I disagree with your first stage configuration. Using a LED as the reference for T3 means that about 5.5mA will flow through R7 and R6. The voltage across R6 if it is 40k will be 220v (or at least try to be). This is more than the rail voltage meaning there is no voltage left for T3. So we do not have a proper constant current source (CCS). Even 18k is too high for 5.5mA as this will try to have 99 volts across it.

If I were you I would reduce the CCS current to 3mA by replacing R7 with a 390 ohm resistor. Then replace R6 with an 18K resistor. I would also replace R4 and R12 with 390 ohm resistors and drop the voltage across these to 0.6 volts (you currently have 1.3 volts here).

In the second stage CCS your choice of 47 ohms as R19 means that 25mA will flow through T7 and T8. This amouns to a dissipation of 4 watts and is completely unnecessary. I would change R19 and R15 to 100ohms and reduce the current through this stage to 12mA.

Cheers
Q
Click to expand...

Hello Quasi,

Have changed and add some features any inprovements welcome
 

Attachments

#3,491
engel dela pena said:
QUASI,
GOOD DAY SiR,
i love the lay out of nmos 400
is a flat pack lay out possible not TO3
Click to expand...

A flat pack version is possible with modification to the output stage tracks and heatsink bracket. Alternatively you could build the Nmos 200 TO247 version and extend the PCB to take additional mosfets. The Nmos350 and Nmos500 is the most compact design for flatpacks (TO247) however.

Cheers
Q
 
#3,493
quasi said:
Hi Gaborbela,

The DC detect circuit does not make any difference to the sound of the amplifier. The detection circuit's input impedance of more than 33K has no effect across the amplifier output and the relay contact has no effect provided it is of good quality. If anyone tells you it does, smile and change the subject.

Cheers
Quasi
Click to expand...

Hello Quasi

In my views relay contact get oxidation and can burning up with high current, a crowbar is more safe for speaker and normally with lower component cost.

In our commercial amplifier we dont use relays for DC Protection and turn on delay/fast off

In future i want try to design 2 Step Class H NMOS


greetings

attachment: latest QSA NMOS PCB design including Input LDR Limiter and Balanced Input, VI Limiter, Crowbar DC Protect
 

Attachments

Last edited:
#3,496
quasi said:
A flat pack version is possible with modification to the output stage tracks and heatsink bracket. Alternatively you could build the Nmos 200 TO247 version and extend the PCB to take additional mosfets. The Nmos350 and Nmos500 is the most compact design for flatpacks (TO247) however.

Cheers
Q
Click to expand...
thanks quasi
i see the nmos 200 extending the output
but i see some resistor changes
i need you confirmation
what resistor should i need to change
regards
 
#3,499
engel dela pena said:
thanks quasi
i see the nmos 200 extending the output
but i see some resistor changes
i need you confirmation
what resistor should i need to change
regards
Click to expand...

The output stage mosfet source resistors should be increased to 0.47 ohms if more than 1 pair of FETs is used.

For rails above 60 volts change R7 to 18k and change R18 to 33k.

Cheers
Q
 
#3,500
riteair said:
Hey Quasi!

I have seen ur Circuit which has following Mistakes in Practical world.

Input Transistors VCE limit is outreached.
The Voltage accross them are more than there specified limit instead of it use 2N5551/2N5401 Trannies.

electrostatic filter
Click to expand...

No. The constant current source ensures that half the voltage (i.e. negative rail to ground) is across R7, T4, R6 and R5/R10 (Nmos350 schematic). This means that the voltage across the input pair is equal to the positive supply rail minus the voltages across R4/R8 and T2/T3.

Regards
Q
 
You must log in or register to reply here.
Top Bottom