Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter

what does that even mean? do I have to keep reminding what application we are talking about? the piezo electric effect is a real effect, but we are not talking about high slewing designs, or rocketships

besides x7r is not bad in this area, its when you start to get to lower grade ceramics. x7r is actually preferred to c0g in many audio applications due to being slightly lossy/less resonant.

we are not placing caps, or resistors, we are placing lumped elements of capacitance, resistance and inductance, especially at high speeds.
 
Last edited:

TNT

Member
Joined 2003
Paid Member
I'll end this conversation with you qusp by replying to another of Your questions and I quote; "marce does PCB design for military applications sometimes, whats your qualification?"

I play the drums, sing in a choir and has served as reference listener during several years in at least 3 different serious high-end development projects. These I actively avoid to name which You will object to but it won't change the situation.

/
 
Hi Guys,

I'm currently on vacation and out of town. I'll try to provide support as possible as I could.

A guy got problem with TPA trident reg working with si570 board, he measure (5V input) the output was just 2.2V rather than 3.3V. I notice somebody else has no problem with this issue. But he may did some mod.

Any help to address this problem? Thank you.

Regards,

Ian
 
hmm, isnt the trident set for 3v3 by default? trident does need a bit more than 5v to work happily in my experience, more happy with 5v5 (for 3v3-3v5vout), but I only have experience with the one on the BII, not the separate tridents sold now. the max load current for the trident is 100mA, which is not quite enough for the Si570 modules 150mA
 
Last edited:
Hi Guys,

I'm currently on vacation and out of town. I'll try to provide support as possible as I could.

A guy got problem with TPA trident reg working with si570 board, he measure (5V input) the output was just 2.2V rather than 3.3V. I notice somebody else has no problem with this issue. But he may did some mod.

Any help to address this problem? Thank you.

Regards,

Ian

May be two reasons : too low ccs current ( should be at least 200mA ) and too low Vin ( should be at least 5.25V). Trident without mod have 80mA CCS.
 
Regarding the 3.3V Trident - it is a shunt regulator. It is very likely just the CCS is too low for your application (not certain how much you need). They come from us configured for use with the Buffalo DAC.

As long as it is a V3 Trident you *should* be able safely increase it to 200ma or so at 5 - 6V input. 5V is usually just fine. You will need to check this for yourself only because 150ma is as far as I have pushed mine - so that is the maximum we state on the site. If it is a V1 I would not try it.

Just monitor the temps - if you see > 65C at the transistors then it is too hot.

Reduce R4 until the desired current is reached. Adding a parallel resistor works great. You can even rig a suitable rheostat connected pot and adjust as needed, but I would replace the pot with a resistor once I know the exact value wanted.

Remember when testing the trident you should provide a dummy load so as not to over-stress the shunt transistor which will otherwise have to shunt the entire CCS current at 3.3v drop (which would not be the case when it is installed in the application).

Also the best bet for getting answers about TPA modules is on our forum. :) Fortunately someone happened to point me here.

Anyway glad to help. Good luck.

Cheers!
Russ
 
Russ, you should change the copy on your website, as I dont have V3, I just repeated the MAX load current stated on the trident webpage.... so its not my misinformation. V1 also does not perform optimally with 5v, better with 5.25-5.5v in my experience.

perhaps you are playing it safe with the general recommendation? fair enough. its a tiny SMD shunt, I didnt expect adding 50% on top of the max load spec would be wise.
 
Last edited:
How should I connect the QNKTC USB-I2S board to this kit? Q N K T C USB-I2S Module and Analog Board

@iancanada, do you have time to check that out?

I think you want to take the signals with the 'DA_' prefix plus the mclk out signal. They're all described on the schematic at the qnktc website link that you posted. These will provide the i2s input that you connect to the FIFO. You will need to mount the on board oscillators on the usb-i2s board to provide the 11.2896 and 12.288MHz MCLK that is required by that board.
 
Hi guys,

I'm just trying to figure out what to use to power my fifo-isolator-si570 combo, and I definitely could use some advice from the experts here.

I’m using TPA Placids in my build, and the one powering the BIIIES heats up to 60-64 C drawing 460 mA, so I’m a little concerned about the heat that another shunted reg would produce with around 650 mA (according to Ian’s docs, that’s what the fifo and the Si570 need). Besides, I’m not sure if a placid can go up to 650 mA, probably not without any mods (bigger heatsinks for sure).

Do you guys think that a sigma11 could be a suitable option?

I’ve been looking trough the thread and seems like some of you are using Salas shunted reg and others are using batteries. Well… here’s where I have to confess my complete ignorance…:eek: can anyone help me a little bit with the passive battery management board? I would really appreciate if someone could point me to an appropriate set of batteries (LiFePo4 maybe?) and a suitable battery charger. What do you use as control voltage? (the fifo's 5V output?)

Many thanks for your help! :)