So the assumption is that the Nano has the boot;oader already installed? which is a good thing for first timers.
I looked at Valery's Protection-01 pcb files that I have on file, its a monster imo, including transformer and all, full featured, well done, I miss the guy, he was a really nice person to converse with, great DIYAudio contributor.
I wait to see it all fit and operate in a chassis, good luck everyone.
I looked at Valery's Protection-01 pcb files that I have on file, its a monster imo, including transformer and all, full featured, well done, I miss the guy, he was a really nice person to converse with, great DIYAudio contributor.
I wait to see it all fit and operate in a chassis, good luck everyone.
I have also looked into adding MCU to my builds to control various functions and protections. STM32 looks promising. Phil's Lab channel on youtube has a lot of tutorials for those. Also, can use this breakout boards, for example, blue pill STM32 board.
I built that , It was kind of big. I built my version of it with the 4$ UNO mini (CH340).
Unless you need the integrated soft start , just pull the 12V/5V off a rail. No need for a separate trafo.
OS
Another cool feature would be remote triggering. Since my amp will be both a stereo and sub amp,
I'd like the presence of audio to trigger it. Here I would need a separate auxiliary trafo , unfortunately.
OS
I'd like the presence of audio to trigger it. Here I would need a separate auxiliary trafo , unfortunately.
OS
Seems kind of a waste to use a 32-bit mcu for such a simple function as protection logic but if they are cheap enough and easy to program might make sense. So many choices today.
I do have another idea about remotely controlled power amplifiers, make the function part of the preamp or controller, have it switch the AC on/off for the amplifiers instead, it needs that function anyways unless you want to get off the couch to turn devices on or off or adjust volume, select sources etc. I have written most of the code for a remotely controlled preamp when I designed my portable radio/media player now 10 years ago, I used a xmega mcu which is a very powerful device for these sort of applications. Iirc I left it at where I had the IR remote working and even designed a hand held IR remote for it.
Not sure if Fairchild still makes a small chip that runs off AC line. It has a 5VDC output at low current for these sort of apps and not require a transformer. Lots to ponder
I do have another idea about remotely controlled power amplifiers, make the function part of the preamp or controller, have it switch the AC on/off for the amplifiers instead, it needs that function anyways unless you want to get off the couch to turn devices on or off or adjust volume, select sources etc. I have written most of the code for a remotely controlled preamp when I designed my portable radio/media player now 10 years ago, I used a xmega mcu which is a very powerful device for these sort of applications. Iirc I left it at where I had the IR remote working and even designed a hand held IR remote for it.
Not sure if Fairchild still makes a small chip that runs off AC line. It has a 5VDC output at low current for these sort of apps and not require a transformer. Lots to ponder
Last edited:
I don't use a preamp , PC controlled ASUS U7 mk2. I have not even seen a CD or LP in 25 years. Movies and TV also
stream through the ASUS.
OS
I leave my bc-1 amp on all the time. In the loft I have an old Yamaha Rx-900u, it’s first generation remote. Many options and features to consider. I still listen to FM too. Last thing I’d be using is a paid for audio streaming service, that’s for sure. Everyone has there requirements, many differ, need h/w to match requirements, fun wow, one thing to keep yourself happy another to keep others happy
Last edited:
Looks like almost everything we designed the amp control systems with is available other than the ULN2803 we used to drive the relay coils with. We had ditched the Nanos and put the microcontroller on board to save space, but I can work with a Nano. The web site is down right now so there's no easy way to see what we put together. Our domains were purchased in Russia so I can't renew them from here with all the sanctions.
We had lots of options to work with the system. There was temperature sensing, remote start input, over-current sensing, voltage loss detection and DC offset detection. The speaker relays were solid state. We designed the DC detection and speaker relays to be modules that hung on the binding posts. We interfaced the speaker relays and the DC detection circuits with opto-isolators too. Analog from the amplifier never mixed directly with digital from out control system. Grounds were isolated completely. Temperature sensing was done with I2C temperature sensors. We has on-board soft-start NTC resistors as well. Over-current sensing was done with opto-isolators too. Current limit could be tailored to the output devices.
As add ons we had rail switches built into our power supplies. We also had an input adapter board that could switch between XLR or RCA input. The control board would check for a switch selection on power up and set the appropriate input before engaging the speakers. We added a small port expander and sent simple binary codes to it for interfacing with a front panel display.
We also put together a front panel display with a RTC and temperature inputs. It would display time when the amp is off, then display all the modes as the amplifier powered up, then display the temperatures of each channel while the amp was running. It was set up as the master temperature monitor and it would signal an interrupt on the control board for overheating. I plan to ditch the current front panel display and switch to ESP32 brains on it when time permits. No need for dinosaur tech like RTC then.
Our systems used on-board transformers but that's not needed. Another option is an AC-DC converter. They're small and cheap. I'm not sure how much noise they put out though.
We had lots of options to work with the system. There was temperature sensing, remote start input, over-current sensing, voltage loss detection and DC offset detection. The speaker relays were solid state. We designed the DC detection and speaker relays to be modules that hung on the binding posts. We interfaced the speaker relays and the DC detection circuits with opto-isolators too. Analog from the amplifier never mixed directly with digital from out control system. Grounds were isolated completely. Temperature sensing was done with I2C temperature sensors. We has on-board soft-start NTC resistors as well. Over-current sensing was done with opto-isolators too. Current limit could be tailored to the output devices.
As add ons we had rail switches built into our power supplies. We also had an input adapter board that could switch between XLR or RCA input. The control board would check for a switch selection on power up and set the appropriate input before engaging the speakers. We added a small port expander and sent simple binary codes to it for interfacing with a front panel display.
We also put together a front panel display with a RTC and temperature inputs. It would display time when the amp is off, then display all the modes as the amplifier powered up, then display the temperatures of each channel while the amp was running. It was set up as the master temperature monitor and it would signal an interrupt on the control board for overheating. I plan to ditch the current front panel display and switch to ESP32 brains on it when time permits. No need for dinosaur tech like RTC then.
Our systems used on-board transformers but that's not needed. Another option is an AC-DC converter. They're small and cheap. I'm not sure how much noise they put out though.
Yes , just the output Re sensor connects to my boards ( or the Wolverine) - the "protect" 3- pin jack.
Wolverine builders are using that jack on their builds.
Found some good replacement semi's for spooky (and other IPS's).
Toshiba has dual HN1C01F and HN1A01F - https://www.mouser.com/datasheet/2/408/HN1C01F_datasheet_en_20210625-1628505.pdf
Same package and pinout as BC807/817 LTP pairs.
The Hellraiser can take the Toshiba NPN/PNP version ( but P/N is reversed) - https://www.mouser.com/datasheet/2/408/HN1B01F_datasheet_en_20210625-2887559.pdf Just flip the SSOT-6 chip around , good to go.
Good quality for $0.40 semi's .....
These are better than the BCxxx's , more linear and made for high end audio. I've seen these used in high end DAC buffer amps.
Does not say they are matched , but diffused on the same die. This would make them much closer than a through-hole device.
I'm working on the BOM's now , the PCB pads are perfect for what I'm choosing. I chose the 1mm pitch (larger) package for all these duals.
NO SOT363's !
I even found cheap 10uF/25V MLCC caps for the regulators. 4.3 X 3.2mm ($0.86) goodness - Kemet - https://www.mouser.com/ProductDetail/KEMET/C1812C106M3RACTU?qs=dB0LedT%2BmmCyXUyd5lGK2g==
Other 10uF's go for over 3$ , screw that !
The only other semi I need to research is what dual op-amp IC I will use for the Spooky and Hellraiser. Lowest current , 24V operation ,
and low cost is what I'm looking for.
I think I can keep my target cost at around 15$ per IPS and still use premium parts !
OS
Last edited:
I have streaming TV. But my movies and music are all top quality 24bit FLAC and 5.1 dolby. I don't like commercials (and $hiity quality) , so
I torrent HD - whole season collections of binge content. The hacked stuff is all at 5.1 and 1080P , almost TOO much dynamic range for the 5.1. Sometimes I'll use DSP compressor post processing to keep others from complaining.
OS
I kind of expected sanctions to affect that ! PM me with any "stuff" you have. I definitely want to include protection on this run of designs.
OS
I would of finished the Spooky BOM ....
But , there are people that will use either the single 14 connector or the 2X14, Should I include both connectors ?
Wurth has cheaper ones - https://www.mouser.com/ProductDetail/Wurth-Elektronik/61301421821?qs=iLbezkQI%2Bsi%2BdVj8CHw1ww==
Have both 1 row and 2 row - only $0.69 !
LED's are another issue , the models I have DO NOT match the datasheets. Most red LED's rated 2V Vf have a Vf of only 1.7-1,8V at the low
current (1.5mA) that I designed for. I designed at this lower current as the I/Vf curve (at 1.5mA) allows this circuit to be servo'ed well.
If I ran at 5mA , I not only would have a much hotter IPS , but the servo op-amp would have to use much more current to shift the Vf
to balance the CCS's. The reds would also blind you at >5mA.
So , depending on what LED - https://www.mouser.com/ProductDetail/Wurth-Elektronik/150120RS75000?qs=2kOmHSv6VfSstSh2FOQbDw==
I would need to see what Vf it would have @ 1.5mA. That wurth LED shows about 1.7-1.8Vf at 1-2mA , according to its datasheet. The current sources
would need 330-390R @ 1.75V and up to 620R with a 2V diode (R14,R21 - spooky).
Should I just include 3 pairs of resistors. ? they are only $0.10 each ?
Would be nice if I could test one with a 6.8K resistor at 12V.
Resistor's, caps ,and semi's are done - with good ,cheap choices. I estimate @ $13 to populate a spook.
BTW - the BOM will be in the spooky package when I'm done.
OS
But , there are people that will use either the single 14 connector or the 2X14, Should I include both connectors ?
Wurth has cheaper ones - https://www.mouser.com/ProductDetail/Wurth-Elektronik/61301421821?qs=iLbezkQI%2Bsi%2BdVj8CHw1ww==
Have both 1 row and 2 row - only $0.69 !
LED's are another issue , the models I have DO NOT match the datasheets. Most red LED's rated 2V Vf have a Vf of only 1.7-1,8V at the low
current (1.5mA) that I designed for. I designed at this lower current as the I/Vf curve (at 1.5mA) allows this circuit to be servo'ed well.
If I ran at 5mA , I not only would have a much hotter IPS , but the servo op-amp would have to use much more current to shift the Vf
to balance the CCS's. The reds would also blind you at >5mA.
So , depending on what LED - https://www.mouser.com/ProductDetail/Wurth-Elektronik/150120RS75000?qs=2kOmHSv6VfSstSh2FOQbDw==
I would need to see what Vf it would have @ 1.5mA. That wurth LED shows about 1.7-1.8Vf at 1-2mA , according to its datasheet. The current sources
would need 330-390R @ 1.75V and up to 620R with a 2V diode (R14,R21 - spooky).
Should I just include 3 pairs of resistors. ? they are only $0.10 each ?
Would be nice if I could test one with a 6.8K resistor at 12V.
Resistor's, caps ,and semi's are done - with good ,cheap choices. I estimate @ $13 to populate a spook.
BTW - the BOM will be in the spooky package when I'm done.
OS
Attachments
Pete,please read this before post the final bom.
It is about,thin & thick film resistors.
Thin film resistors recommended for audio amplications.
It is about,thin & thick film resistors.
Thin film resistors recommended for audio amplications.
Last edited:
Thank you , my friend. The paper is kind of vague and subjective - $$$ .... a sales doc.
Kind of like the perceived effect of a 15$ audiophile input cap.
Thin film would be good as a (potted) lab grade reference.
My design could outperform most with even 5% resistors , it's servo'ed with 100db of NFB.
Thick film (what I picked) , is better for humid environments + reliability .... and cheap.
The effect of Tc on the Hfe of the transistors or the Vf of the LED's will be 100X the difference of thin vs. thick film.
PS - your typical 1% metal film through-hole resistor is thick film. SMD thick film is the same ceramic (square block) with the vapor deposit.
OS
Thin film resistors are supposed to be equivalent to the metal film through hole resistors we were accustomed to using. Thick film work and measure okay, but the act like fuses in case of short. The blow open easily which can be a pain to diagnose. Yaego has some inexpensive thin film resistor that work well.
Last edited:
You can simulate this....
Wurth is the only LED manufacturer I know of that supplies LTspice models. LTspice now include these LED models as a standard library.
To add the Wurth LED to the schematic:
goto >edit>component>[Contrib]>[Wurth]>[Optoelectronics]>[SignalLED]>Mono-color]>WL-SMCW, then add to schematic.
On the schematic, right click mouse on the LED to bring up the properties, double click on "SPICEMODEL" and a drop down menu should be available, find the 1206_150120RS75000 LED in the list then click OK, the other LED you can use is the green, its part number is 1206_150120GS75000.
I found a Red LED from Kingbright Vf 1.8v @ 2mA - https://mouser.com/ProductDetail/Kingbright/APT3216LSECK-J3-PRV
a green version is Vf 2.65v @ 2mA - https://mouser.com/ProductDetail/Kingbright/APT3216LZGCK
I don't think there will be any noticeable degradation with thick film. As far as diagnosis's - just uplug the IPS , throw in garbage ... replace !
At $13 , why bother fixing it ? I doubt many will have any fail , my spook is 8 years old , no issue.
This makes these IPS's like a network or cheap sound card in a PC , just swap in a new one.
I'll still price out the thin film's , if they are under .20C .... I'll use them.
Mouser is down for maintainence. The included LT Qtlp690 does 1.75 at 1.5ma and 1.9 at 5ma , this matches the Wurth datasheet curve exactly.
Since the LED CCS is the only factor that really matters in this design , real world testing on one might be the only way.
BTW , a 2Q CCS would also have to have Re tweaked as Vbe would vary with different semi's.
OS