Arduino+PGA2311 Ultimate PreAmp with OLED and IO switching

[zdr]

It's finally finished. Nothing spectacular, but feature rich, pretty damn good sounding and fancy lookin'. Driven by a single, cheap 7.5-9V switching or linear power supply, with 5 inputs and 1 output with tape monitor , switched by small signal relays. Controlled by Apple TV compatible remote, panel operable with two independent encoders, with graphic 2.4" OLED. Attenuation done with PGA2311 running on +-5V with negative rail inverted from +5V and stabilized with -5V LDO. Firmware ported from LDR Pre MkII. Volume steps are 1dB, from 0-99dB

Why PGA2311? Because I heard Musical Fidelity M3Si which is using the similar chip for volume control, and it sounded fantastic.

All passives are SMD 1206, easy to solder even for 56 year olds like me.

Some of the features:

1. Sleep timer (1 - 255 minutes, 0 for disable). Also switches off external relay if connected, for turning off power amp for example.
2. Screen minimum brightness (0-255)
3. Screen maximum brightness (0-255). Screen dims down from max to min after Screen Saver timeout
4. Screen saver timeout
5. Channel Balance (+-20dB between channels)
6. Fully customizable input and output names
7. Fully customizable welcome message
8. Screen off time (1-255 minutes, 0 for disable)
9. Safe volume level - volume value will not be saved on power down above this value, to prevent a surprise on next power up.
10. Child lock - disable encoders and hide a remote
11. Automatic delayed power amp control, also manually switchable.

Operation: PGA preamp mki - YouTube
https://sites.google.com/view/dacgear/pga-pre

 

[Montgomery HIFI]

Oh man that is so cool! Does the Arduino DSP the input to switch to the different outputs? Could this be turned into a DAC

 

[zdr]

Oh man that is so cool! Does the Arduino DSP the input to switch to the different outputs? Could this be turned into a DAC

There is no DSP capability in Arduino, it's just too slow for that (afaik). So no DAC, just volume control and IO switching.

 

[garym999]

Hi zdr,


You may recall we exchanged a few posts way back on the LDR MK1 pre-amp. I then spent ages with the LDRs looking a temp compensation and calibration etc. eventually running out of steam and sticking with my MK1 build. Also I was a little put off using LDRs and was leaning towards a resistor/relay ladder but this might be a nice stop gap.



I've been wanting to revisit for a while and seeing that you have not stood still since and produced this I'm inclined to look at it again.


I see the bare board is available from your website what about the BOM and codebase?

 

[zdr]

Firmware is not open source nor is the PCB layout, if that's what you are asking. Schematics is posted above, and I will try to make the BOM out of it soon. All passive parts are SMD 1206, 1% resistors, all caps X7R except one big elco which is 1000-2200uF 6.3V. Relays are the same as in LDR Pre MkII, non latching:

AGN2004H Panasonic Industrial Devices | Mouser Belgique

SMD or through hole, they both fit on PCB. Any relay with the same footprint for 5V or 4.5V should work.

This is the voltage inverter:
TPS60402DBVR Texas Instruments | Mouser Belgique

LDO regulator for positive voltage rail:
TLV702475DBVR Texas Instruments | Mouser Belgique

LDO regulator for negative rail:
TPS72301DBVR Texas Instruments | Mouser Belgique

four smd 7805 regulators:
https://www.mouser.be/ProductDetail...ha2pyFadujoaRTxonov%2Blqs4EwRSRHCOt5nWya0nZA=

PGA2311:
https://www.mouser.be/ProductDetail/Texas-Instruments/PGA2311UA?qs=68g8TF7PocK5AEXipt0WLg==

SMD caps:
https://www.mouser.be/ProductDetail/KEMET/C1206Y104K9RACTU?qs=rUPLZ6nenJPIO9Oy/dArjw==

 

[454Casull]

Does it have the ability to expand for control of additional channels?

 

[zdr]

Does it have the ability to expand for control of additional channels?

PGA2311 has the ability to daisy chain, but I did not connect SDO pin to any of the external connectors. That's the easy part, scaling up the relays would complicate things fast. I am not a big fan of multichannel audio though.

 

[454Casull]

I am not a big fan of multichannel audio though.

That's one of the biggest advantages of the PGA2311 though...

 

[zdr]

That's one of the biggest advantages of the PGA2311 though...

Not in this case really. Daisy chaining would have to be done on Arduino level anyway to multiply the channel switching as well, in which case we don't really care who is executing downstream volume control commands.

 

[zdr]

Jumper that needs to be placed:

 

[mercedese63]

Why PGA2311? Because I heard Musical Fidelity M3Si which is using the similar chip for volume control, and it sounded fantastic.

The T.I. PGA2311 totally BLOWS and is simply NO match for the MUSES 72320 when it comes to SQ.

LOL...Musical Fidelity is now just one of many companies who's having their garbage products manufactured to a low price point in CHINA.
Musical Fidelity's BEST products, which were designed by Antony Michaelson and manufactured in England years ago, NEVER used the crappy-sounding PGA2311 chip!

 

[jean-paul]

Hi the better performing PGA2310 would have been a better choice when sound quality is important. That one will give sufficient headroom too. You just need to up the analog voltages. Design flaw is the omitted input buffer as both IC's need one for sure! The buffer needs to be AC coupled and it needs an RC input filter (good design practice). Second flaw is the too low +/- 4.75V in an already tight scenario regarding headroom. 4.75V is the absolute minimum too for functioning of the chip. IF you stick with PGA2311 you really need to use +/- 5V.

Third flaw is the TPS60401 as its internal switching frequency is between 13 and 20 kHz which is exactly the area we don't want and certainly not when it is used for the negative supply voltage of the analog side of things.

The design would benefit from a symmetric power supply and preferably a small Rcore transformer. For a slightly higher cost way better performance... Again that is only valid if sound quality is important as there seem to be more and more designs introduced that don't care about sound quality but focus on easy building or features.

PGA2310/2311 are not the cream of the crop but they can perform OK when you stick to the requirements.

 

[zdr]

Do you have data to backup the “better performing” claim? Because I have data to contradict it right here:

https://e2e.ti.com/cfs-file/__key/c...GA4311_2C00_-PGA2310-THD_2B00_N-vs.-Gain.xlsx

 

[jean-paul]

It depends what you call "better performing" PGA2310 possible 4V rms output gives it away. PGA2311 has 2.5V rms absolute maximum for input voltage before clipping. That is tight with modern sources. Regarding sources: let's say any modern source that has +/- 5V supply voltages in its output stage outputting 2V rms will be shot on sight

BTW in the sheet PGA4311 is mentioned, I don't know that one and if it is completely comparable to PGA2311. The sheet also uses the highly unrealistic load of 100 kOhm.

Hey, it is your call. I just mention my experiences with the IC's to prevent the usual pitfalls. PGA2310/2311 benefit tremendously from both input- and output buffers. In fact, I think they can not be used seriously without buffers which makes them less attractive as designs are more complicated and a higher part count is necessary.

MUSES72320 performs way better than both the PGA chips. It also needs a symmetric PSU higher than +/- 8.5V and it is not easily "overdriven". It does not need input buffering BUT it definitely needs output buffering.

 

[jean-paul]

Wild guess: PGA2311 will not be able to cope with 2.5V rms input/output voltages when used with +/- 4.75V. I tried to get this out of the datasheet but failed to find it as TI is not clear on this. They state that with analog voltages of +/- 5V both input and output can be 7.5 Vpp. This would mean 7 Vpp with +/- 4.75V supply voltages. TI says the chip can have 2.5 V rms when used with +/- 5V supply voltages but this does not comply with 7.5 Vpp....

Safe guess is slightly above 2V rms as a maximum. Too close to call it "enough headroom" for many designers. Again, critical remarks are meant constructively.

 

[zdr]

Why are we even talking about 2Vrms here, am I missing something? Consumer audio standard for line level is 0.316 Vrms and professional audio 1.228:

Line level - Wikipedia

PGA4311 is just two 2311 in one.

 

[jean-paul]

That seems very old information. CD players etc. are 2 Vrms (I think +/- 0.3 V even) since they were invented in the eighties. Mediaplayers also use 2 Vrms as a standard. I am not sure about the high and mighty source the smart phone as I refuse to use one for audio. Meanwhile the 2 Vrms is standard. Many DAB+ tuners are specified for 2.1 Vrms. Bleutooth receivers of some quality also use 2 Vrms. WiFi audio receivers mostly use 2 Vrms. So except (?) for the smart phone almost ALL modern "quality" sources with signal ended outputs are 2 Vrms... with which the ultimate preamp should be able to cope.

This line level differences in the past always were a nuisance when one had, for instance, an old but very good FM tuner adhering to the older standards. Switching sources could have a surprise effect. But I see you choose not to want constructive remarks. So be it, have fun with the project.

 

[zdr]

That seems very old information. CD players etc. are 2 Vrms (I think +/- 0.3 V even) since they were invented in the eighties. Mediaplayers also use 2 Vrms as a standard. I am not sure about the high and mighty source the smart phone as I refuse to use one for audio. Meanwhile the 2 Vrms is standard. Many DAB+ tuners are specified for 2.1 Vrms. Bleutooth receivers of some quality also use 2 Vrms. WiFi audio receivers mostly use 2 Vrms. So except (?) for the smart phone almost ALL modern sources with signal ended outputs are 2 Vrms...

This line level differences in the past always were a nuisance when one had, for instance, an old but very good FM tuner adhering to the older standards. Switching sources could have a surprise effect. But I see you choose not to want constructive remarks. So be it, have fun with the project.

I keep posting links and data here, you keep disputing them without backing up your claims, yet you see yourself as constructive - ok

I have two DENON SACD players from 2004, one Cambridge Audio CXN v1 network streamer from around 2016, Topping D50 DAC and my XR+DAC which is also built with the same output level as PGA PreAmp. They all play on the (audibly) same level. I would be very annoyed to have to switch between sources that differ from each other as much as 20dB. I agree with one thing though, I don't consider smart phone as a hifi source.

Please post the links to "new" standard, until then I will keep complying with the "old" one.

What is “Line” Level and What is “Mic” Level? | Where It’s A-T).

 

[zdr]

I have posted a new firmware, v1.21. It has a sleep timer built in, default value 240 min, configurable from 0 (disable) to 255, from setup menu. After this time, device goes in quasi standby mode, turning the OLED and relays off.

 

[jean-paul]

Please post the links to "new" standard, until then I will keep complying with the "old" one.

Link showing real world output level of your Topping DAC:

Review and Measurements of New Topping D50s DAC | Audio Science Review (ASR) Forum

The CXN specifications are hard to find but I am sure it also has 2Vrms line outputs as all recent gear does. If you look up the unspecified Denon SACD players you will definitely find the same. I am pretty sure you know how to use google yourself so please find out about the "new" standard

Some DAC chip datasheets mention it clearly:

https://statics.cirrus.com/pubs/proDatasheet/CS4354_F3.pdf

It is of course the maximum output line level and not average but we design with maximum and/or worst case numbers in general.