UcD project, Subwoofer class d amp, full h-bridge, single supply +50 V

[lorylaci]

ok then, i will use IRF540N for testing, now i need to search for HIP ics. If its success i`ll use IRFB4710 its good enough i think

IRFB4710 has very high gate charge (even the HIP4080A'a 2,5A is not enough for that), and a very bad diode. Look at IRF540Z datasheet, it has a 62nC max diode, pretty good (and 63nC gate charge).
Bad diode not only means losses, but will give you high EMI, which will cause noise and distortion also.

Check Intersil's website first. Then check the distributors for your country. HIP4080A is not so expensive (its cheaper than an IRS2092).

 

[norazmi]

got it, im order directly from intersil and request sample too.

 

[lorylaci]

got it, im order directly from intersil and request sample too.

If you order samples from Intersil (I did too), don't order too much things, otherwise you will have to pay customes (usually up to 5 pieces of ICs it is below customs value)
(there was one ocassion when I had to pay customs )

 

[norazmi]

i just order 10 pcs and + 2 samples, here in my country tax will be charge if order more total amount more than $200 usd. I think i`ll passed it

 

[emilvv]

Hello ..

Hello,

I have a new question about HIP4081 ...

As lorylaci suggested here: ....

http://www.diyaudio.com/forums/clas...-bridge-single-supply-50-v-2.html#post2934047

Can you tell me how max output power can be released from: ...

http://www.diyaudio.com/forums/class-d/132081-single-supply-bridge-ucd-referencing-signals-5.html

see .... 400w single supply btl ucd.pdf ..

and how much will be THD+N with 56 V DC and 8 ohm load ( and 4 ohm load to ... ) ??

lorylaci says : ... "Because IR2110 has limited pulse width" ... so is the HIP4081 better ???

Also .... "But using HIP4080 with integarated high-speed comparator can raise your modulation limit" .... so ... using HIP4081 + external TL3016 is better than combination of LT1016 and 2 x IR2110 ??

All answers are for subwoofer amp ... 0-300 Hz ... 8/4 ohm load ..

Regards,

Emil

 

[lorylaci]

Hello,

I have a new question about HIP4081 ...

As lorylaci suggested here: ....

http://www.diyaudio.com/forums/clas...-bridge-single-supply-50-v-2.html#post2934047

Can you tell me how max output power can be released from: ...

http://www.diyaudio.com/forums/class-d/132081-single-supply-bridge-ucd-referencing-signals-5.html

see .... 400w single supply btl ucd.pdf ..

and how much will be THD+N with 56 V DC and 8 ohm load ( and 4 ohm load to ... ) ??

lorylaci says : ... "Because IR2110 has limited pulse width" ... so is the HIP4081 better ???

Also .... "But using HIP4080 with integarated high-speed comparator can raise your modulation limit" .... so ... using HIP4081 + external TL3016 is better than combination of LT1016 and 2 x IR2110 ??

All answers are for subwoofer amp ... 0-300 Hz ... 8/4 ohm load ..

Regards,

Emil

Emil, please take the pain to read the datasheets, and don't ask what is obvious. HIP408x has a maximal supply limit of 80V, sok you can calculate the max output power to all loads from that.

HIP408x has better gate driving, dead time setting ability, and no minimal pulse width error (or at least I haven't head about it). Using HIP4080 is better than HIP4081+external comparator. Using HIP4081 with comparator, is shure better than using IR2110s. IR2110 has much higher prop delay.

CroMagnoic did not state any THD+N measuremns, so I cannot say that either. Why don't you ask him? But this model with good dead-time adjustment is propapbly capable of THD+N < 0,1%.

 

[norazmi]

Lorylaci, i`ve receive Hip ics, BUT its soic package , damn, i`m dead with soic arghhhh....

 

[hanshuang]

Schematic is different from pcb

Hi lorylaci,

I found the C2 connect C16,C17 on the schematic, but the C2 connect C14 and C15 on the PCB. Which is correct? thanks!

 

[emilvv]

unbalanced to balanced sch

Hello,

Is it a good unbalanced to balanced schematic ?

BR

 

[Pafi]

Acceptable. +out has less phase-shift at high freq than -out.

 

[Eva]

The TL074 has JFET inputs whose bias current is in the order of nanoamperes, and whose bandwidth does not extend to many Mhz, the 2.2k series resistors at IN+ are redundant. The 100k,100k divider can be changed to 10k+10k for less phase shift.

The vulnerability of the TL074 is output impedance which becomes negative above a few Mhz, feeding fast current glitches into the output shows the worst of TL074 and similar, same happens with LM324 and similar. 100r series, 1nF to GND is recommended in this scenario. MC33078/MC33079/LM833/LM837 are recommended for audio and similarly priced, and are also classic mature affordable rugged parts. In case of using these parts, low value resistors are recommended (as to feed inputs with impedance below 6k or so for optimum noise) and ~100pF capacitors from output to IN- in inverting configuration are recommended for optimum HF behavior of op-amp, also keeping an eye in the resulting audio roll-off. For applications where the poles of OUT+ and OUT- should be matched a RC is recommended at the IN+ of the OUT+ op-amp.

The tradition of using polar electrolytics for audio in circuits with no bias is something to be revised, as these capacitors are non-linearly lossy used that way. I recommend SMD ceramics, which are available in values up to 47uF 10V in 1206 size, these exhibit the best linearity around zero bias and >10x longer life.