@mason_f8 @Mooly @abraxalito @tomchr
thanks for your comments!
Mouser's order arrived today with lots of candies to play with.
I have designed a schematic with the idea I have in mind.
I want to make a PCB with 3 different amplifiers (OPA1622, OPA1688 and TPA6130A2) to test each one individually and thus decide which one I am going to use in production.
I am going to test the LM27761 to generate negative 5v and thus avoid the decoupling capacitors. I also put them in the schematic, and they will be bypassed in the tests.
I have also placed PinHeaders to select the input and output of each chipamp, so as not to mix signals.
What I am not very clear about are the values of the feedback or gain resistors and capacitors of each IC.
I am only going to reproduce frequencies from 250hz to 8000hz
I await your comments, any help is valuable to me
thanks
Leonardo
thanks for your comments!
Mouser's order arrived today with lots of candies to play with.
I have designed a schematic with the idea I have in mind.
I want to make a PCB with 3 different amplifiers (OPA1622, OPA1688 and TPA6130A2) to test each one individually and thus decide which one I am going to use in production.
I am going to test the LM27761 to generate negative 5v and thus avoid the decoupling capacitors. I also put them in the schematic, and they will be bypassed in the tests.
I have also placed PinHeaders to select the input and output of each chipamp, so as not to mix signals.
What I am not very clear about are the values of the feedback or gain resistors and capacitors of each IC.
I am only going to reproduce frequencies from 250hz to 8000hz
I await your comments, any help is valuable to me
thanks
Leonardo
I am a little perplexed. Using ±5V supplies is a good idea. However, the LM27761 input is up to 5V (Absolute Maximum 5.8V). Wasn't it a 9V battery?
XD in a good or bad way?
yes, the battery feeds the 7805 and thus feeds the LM27761
sorry! I gave up with the 2822 circuit,
It has a lot of gain for my requirements.
I'm going to try these other chipamps
Your schematic looks like the one I used for my headamp : AmpCasq. I can share some experience :
LM27761 is super picky regarding pcb layout, you need ground planes and short leads, that seems ok on your project.
Output impedance of the charge pump is related to 4 times the ESR of the flying cap (C10 on your schematic), rather than using a single cap for C10, use 2 in parallel in order to lower ESR.
On heavy load, LM27761 charge pump may suffer to provide enough voltage to the internal LDO. -5V may drop. It could be interesting to gain few hundred of mv on LM27661 input side: try to push up reference of 7805 by 500mv and reach LM27761 max input voltage (5.5V). Use a resistor divider with 470Ω/47Ω.
Sorry, I had overlooked the 7805 in the schematic.
Alkaline batteries may drop in voltage by about 30% before they run out of capacity. Also, the 9V battery has a large internal resistance, so it is possible that the voltage will drop when a large output is produced. 78XX requires at least 2V between in and out. It is recommended to use a low loss type. Even so, it is necessary to check what the power supply voltage will be at maximum output and how long the battery will last.
Since the output of the AD9833 is 0.6V and this value is Vpp, @abraxalito #23, which considers 200mVrms for a sine wave, is reasonable.
A gain G=9 is required to obtain a maximum output of 100mW 32Ω (1.8Vrms), and the sound pressure at that time is 122dB SPL😱
With the OPA1688, only about 1.2Vrms can be obtained with non-clip from Figures 7-18 and 9-3 of DS.
To reduce noise on the output, the shorthand is to use an amplifier with low input noise density and discard excess gain.
@abraxalito's recommendation G=5 (14dB) gives 1V 31mW 112dB SPL.
In that case, if R16, R13, R18, R17 in the schematic are set to 220R, G=14.9dB can be obtained.
DC at the output draws unnecessary current through the HP, which may increase current consumption, if not destroy it. It is recommended to insert C of about 100uF between 220R and GND. When using polarity, the input of OPA1622 is NPN, so please set the GND side to +. Also, for the OPA1622, it is recommended that R9 and R10 be 10kΩ or less.
Getting 100dB of attenuation with a pot is probably not possible.
Even if the L pad is configured using fixed resistors (for example, 1MΩ, 10Ω or 100kΩ, 1Ω), it will be affected by stray capacitance and adjustment will be necessary to flatten the frequency characteristics.
But before that,
You'll need to refer to the PRR#34 post to see if you really need a 10dB SPL test signal.
Last edited:
Yes! I have copied part of your schematic.
Thank you for your advice on the LM27761, I will apply them
@mason_f8
wow!, thank you very much for taking the time to read the whole post!
Is the calculation the same for the 2 chipamps? (1688 and 1622)
It's not db SPL, it's db HL as @PRR rightly said
In any case, the maximum output is around 110db SPL, which varies according to the frequency.
but it affects the minimum levels, since for example, 0db HL at 125hz is equivalent to 45db SPL, at 1000hz it is equivalent to 7.5db SPL and at 8000hz it is equivalent to 15.5db SPL (for reference)
so the minimum would be 10db HL at each frequency, I have to write a conversion table
wow!, thank you very much for taking the time to read the whole post!
how to calculate this? I would like to know
Is the calculation the same for the 2 chipamps? (1688 and 1622)
you say directly in the output? OUT -> C100uf -> R220 -> GND
I do not understand this. how do i connect it?
I spoke with the audiology expert and he corrected the unit of measurement error.
It's not db SPL, it's db HL as @PRR rightly said
In any case, the maximum output is around 110db SPL, which varies according to the frequency.
but it affects the minimum levels, since for example, 0db HL at 125hz is equivalent to 45db SPL, at 1000hz it is equivalent to 7.5db SPL and at 8000hz it is equivalent to 15.5db SPL (for reference)
so the minimum would be 10db HL at each frequency, I have to write a conversion table
Ok, I built the first test amp. with a TPA6130A2.
It is simple 5v power. being able to configure the Gain and the output mode (Stereo, Mono and BTL)
It works very well, there is no noise at all in the headphones.
I have configured it at its maximum volume and in BTL mode to obtain the maximum power at the output. At first glance it sounds great.
but I have not been able to measure yet with the minidsp ears.
I'm going to take measurements tomorrow.
Now I have a query
How could I increase the output power in case I need more?
other.
can the OPA1622 be used in BTL mode?
(the brown stain is from the hot air that has burned the flux)
It is simple 5v power. being able to configure the Gain and the output mode (Stereo, Mono and BTL)
It works very well, there is no noise at all in the headphones.
I have configured it at its maximum volume and in BTL mode to obtain the maximum power at the output. At first glance it sounds great.
but I have not been able to measure yet with the minidsp ears.
I'm going to take measurements tomorrow.
Now I have a query
How could I increase the output power in case I need more?
other.
can the OPA1622 be used in BTL mode?
(the brown stain is from the hot air that has burned the flux)
I have a question - I see a three wire interface between your amp and your headphones (L,R and GND). How is it you can run in BTL? Are you running mono with L and R transducers in series?
Well, I measured with minidsp ears the circuit with the TPA6130A2, the sound is really clean, without hiss or power supply noise.
but unfortunately I couldn't get over 100db SPL@32ohms
it is really very stable at all frequencies, 98-100db SPL
I must reach 120db SPL (100db HL) in some frequencies.
I don't think an OPA1622 alone will far outperform the TPA6130A2.
If I put another amplifier at the output of the TPA6130A2, can it increase its output? which will be ?
My question is, can the OPA1622 be used in bridge mode?
If so, how many OPA1622 would I have to place in parallel to reach that power?
thanks in advance
but unfortunately I couldn't get over 100db SPL@32ohms
it is really very stable at all frequencies, 98-100db SPL
I must reach 120db SPL (100db HL) in some frequencies.
I don't think an OPA1622 alone will far outperform the TPA6130A2.
If I put another amplifier at the output of the TPA6130A2, can it increase its output? which will be ?
My question is, can the OPA1622 be used in bridge mode?
If so, how many OPA1622 would I have to place in parallel to reach that power?
thanks in advance