His symbol for the the detector optoisolator is wrong. If you check the datasheet it's a bidirectional emitter type so it will detect negative DC.
The Delay() functions look like intentional startup delay to allow the amp to settle before engaging the speakers which is pretty much necessary. There's no delays happening once the speaker relays are engaged.
Why do you think it would not detect negative DC! Please explain.
The used optocoupler is a ac type with 2 leds inside. This means it will work with positive and negative DC at the RC highpass filter build with a non polarized electrolytic capacitor.
There is no delay() function in loop() routine as soon it changes state to RUN. It enters RUN state directly after turning speaker on (see line 77: digitalWrite(SpkrON, HIGH))
The StartDelay() function is used to waits 4 seconds before it switches on.
The problem had to do with the optocoupler as already kindly pointed out by @jwilhelm
I haven't looked at your code in 6 months and cant remember where and when the risk occurred, but it might had to do with not latching a DC fault over a power off cycle. Can't be bordered to spend hours finding it again. I had a quick look and there is three occurrences of Delay() one within Loop() its conditional. You also jump to StartDelay() from within Loop(). Imo it would have been better if you polled the system clock, this considering the non responsiveness during Delay(). Should I have misinterpreted the code I offer apologies.
I haven't looked at your code in 6 months and cant remember where and when the risk occurred, but it might had to do with not latching a DC fault over a power off cycle. Can't be bordered to spend hours finding it again. I had a quick look and there is three occurrences of Delay() one within Loop() its conditional. You also jump to StartDelay() from within Loop(). Imo it would have been better if you polled the system clock, this considering the non responsiveness during Delay(). Should I have misinterpreted the code I offer apologies.
I like felix1024's protection circuit and intent to make and test my own version once I find the time and parts. Would it be a good idea to implement a hardwired DC protection by adding a transistor that pulls down the base of Q3? In this case speakers will be DC protected even when the processor is delayed or not working.
well there is no reason why the MCU should stop working at all. (not in several month on my amp) The MCU latches the error state and show the reason of shutdown using the leds until the amp is switched off.
The only glitch I noticed was sometime a spurious reset of the MCU caused by not connecting the RESET pin. Corrected by pull pin 4 to +5V, see actual schematic.
The only glitch I noticed was sometime a spurious reset of the MCU caused by not connecting the RESET pin. Corrected by pull pin 4 to +5V, see actual schematic.
@kokkie I remember @jwilhelm has development a system that does exactly that, but it's considerable more complicated then just one extra transistor.
Self I am working on on a similar system, it will handle DC protect by a hardware interrupt using an ATmega328p, it will also take care of the soft start procedure, vandal switch, overcurrent protection and remote startup by 12V from preamplifier. No ETA.
Self I am working on on a similar system, it will handle DC protect by a hardware interrupt using an ATmega328p, it will also take care of the soft start procedure, vandal switch, overcurrent protection and remote startup by 12V from preamplifier. No ETA.
We went a lot more in depth with our system. It's a complete amplifier housekeeping circuit. One thing I found early on is they can cause terrible ground loops, especially in a stereo system so I took the time to completely isolate the digital and analog systems. I also kept the actual speaker grounds disconnected from the detector circuits to eliminate dual ground paths. Lots of planning and work to pull it off correctly.
@felix1024 you are right that there is no reason to worry, but I would sleep better with a backup protection circuit. @OmeEd, once finished I hope you are planning on sharing you schematic and program ?
Chewing my way through this thread but wanted to order some boards, so may I beg some guidance on these premature questions?
1. Would +/-48V voltage rails be OK into a 4 ohm load; thinking of two (V+ & V-) Meanwell LRS-600N2-48 (rated at 12.5A @48V and up to 5 seconds at 200%)
2. I have built stuff prior (a Bruce Heron tube amp, M2X Pass clone, Wayne's Headphone/line stage, Mark Johnsons VRDN supply), given the high level of support here and many fellow builders on line, is this build `a reach' without a lot of diagnostics?
3. On that note, I have basic hand tools, DVM, soldering iron etc., what additional do I need to build this assuming that I don't need to get into tweaking performance beyond the functional 'out of the box' result?
1. Would +/-48V voltage rails be OK into a 4 ohm load; thinking of two (V+ & V-) Meanwell LRS-600N2-48 (rated at 12.5A @48V and up to 5 seconds at 200%)
2. I have built stuff prior (a Bruce Heron tube amp, M2X Pass clone, Wayne's Headphone/line stage, Mark Johnsons VRDN supply), given the high level of support here and many fellow builders on line, is this build `a reach' without a lot of diagnostics?
3. On that note, I have basic hand tools, DVM, soldering iron etc., what additional do I need to build this assuming that I don't need to get into tweaking performance beyond the functional 'out of the box' result?
1. Lowest supply voltage for SMTP mentioned in the Bill Of Materials is 2 X 55 Volts, this imply's that 48 Volt is not an obvious choice. It can be made to work but part values wont be optimal and may require adjustments.
2. If you follow Stuarts guide and work accurately you will end up with a working amplifier. If you change course into uncharted waters without the necessary knowledge, you could find yourself in an unfortunate situation.
3. No high tech stuff is needed, but some workbench equipment will make life easier. A decent tip on your soldering iron and a choice for a 'workable' solder are important.
2. If you follow Stuarts guide and work accurately you will end up with a working amplifier. If you change course into uncharted waters without the necessary knowledge, you could find yourself in an unfortunate situation.
3. No high tech stuff is needed, but some workbench equipment will make life easier. A decent tip on your soldering iron and a choice for a 'workable' solder are important.
@Nikos maybe you must try a better capacitor first.
Really, all capacitors sound somehow. It is unavoidable.
1. SMPS is available at +/-55V so all good there.
2. No Issue about sticking to the beaten path.
3. Have decent iron and solder/sucker, pcb clamp/vise thing and have played with SMD on a previous project, so I think I am OK there.
Can all optimization be performed with the aid of DVM and no scope required?
2. No Issue about sticking to the beaten path.
3. Have decent iron and solder/sucker, pcb clamp/vise thing and have played with SMD on a previous project, so I think I am OK there.
Can all optimization be performed with the aid of DVM and no scope required?
@OmeEd.
Τhe amplifier is amazing in terms of stability and with a construction manual which is the first time I have seen it in so many years on the forum.But as you know, a generator and an oscilloscope are always necessary when strange things happen and not only. Confirm that you have done what you should....
Last edited:
Physics is not rolling dice. Circuits build on printed circuit boards act very predictable. Wave goodbye to Ebay and Aliexpress, buy parts that stay within the required tolerances and only from reliable sources. Providing you work accurately, all your Wolverines will work. The needed adjustments can be done with a DMM. Oscilloscope is great for faultfinding and development. Providing you know how to use it, what to measure with it, how to measure it, how to interpret the measured results, and what to do if the measured results are not as desired. BTW, accurately pairing the long tailed pair and its degeneration resistors can make quite a difference in the achieved result. Act upfront.