Perhaps a stupid question.. Is there a way to bypass C2 wbst using the 'mixed' feedback layout?
Thanks
Thanks
Thanks Vunce, is that the voltage at the 604's legs? I'm guessing that you don't see any switch on/ off spikes, due to the Rs resistance and the local capMine are fed +/-23V5
You could increase the Rs resistors to drop a bit of voltage feeding opamp.
I'd also like to know why everyone is using large(ish) local psu caps on their boards. Thanks
If anyone is reading this and wants a set of 4 Mills MRA 0R56 5W bias resistors, 4-off 470R 0.5W Audio Note Niobium gate stoppers and a pair of 22pF Charcroft silver mica caps then PM me.
Does anyone have any views / preferences for either of the following as local power decoupling caps:
Pana FM 1500uF 35V, ESR ~8Ohms 3.7A ripple, 12.5mm*35.5mm. £2.32 each
Nichicon UHW 5600uF, 25V, ESR ~8Ohms 4.2A, 16*35.5mm. £3.49 each.
thanks 😁
Pana FM 1500uF 35V, ESR ~8Ohms 3.7A ripple, 12.5mm*35.5mm. £2.32 each
Nichicon UHW 5600uF, 25V, ESR ~8Ohms 4.2A, 16*35.5mm. £3.49 each.
thanks 😁
They won’t fit on the board. Best installed as separate pairs for each channel, on top of or alongside the main PSU board as space demands.
This does beg the question; Is ESR the absolute concern for a power amp OPS local decoupling cap? Ie what is the point of placing huge capacity/ ave ESR caps on the board (Wurth 10,000uF, etc).
To my mind the larger metal cans will pick up more radiated noise from the transformer. Just trying to understand what people were thinking here.
To my mind the larger metal cans will pick up more radiated noise from the transformer. Just trying to understand what people were thinking here.
The question is “what issue is a set of local decoupling caps trying to solve?”
There are different approaches depending on whether one is using a shared supply or dual-mono.
There are different approaches depending on whether one is using a shared supply or dual-mono.
Fixing physically large lytic de-coupling caps to the boards/ legs of output devices is a means for noise injection.
Did anyone else building this amp detect any instability at all with the OPA604 and mixed feedback? Shouldn't be the case
Too much H2 for you Patrick?Some people absolutely love Accuphase.
But not everyone.
But then it is a better world that we have diversity and free choice, not ?
🤓
Patrick
I just got pcbs for the M2OPS from jlcpcb, which looks like a cool thing to try out. It's so easy and cheap to order pcbs now. I have three projects based on Patrick's threads lined up all with boards ready in jlcpcb boxes... Thank you, Patrick!
For this one I have a Pass-style PSU I built for a different M2 build a couple of years ago that I could use, but I'm wondering about doing something different. One of my weaknesses is e**y; in particular I find nice transformers at a terrific price irresistible. I'm like a magpie - see something pretty, gotta have it... "Oooohhh..... shiny!!!" So I recently picked up a pair of Avel Lindberg 160VA transformers (the old ones from the UK) with 25V-0-25V secondaries. AT less than $9 each and modest shipping I couldn't resist. So they're just a little small, but much higher voltage than the usual transformers you'd choose for Pass-style monoblocks. But this discussion about using the M2OPS in Class AB got me thinking...
Suppose I were to build M2OPS monoblocks with these trafos. After the recitifiers I'll have +35V - 0 - 35V before smoothing, so if I estimate I have about 70W to burn (using the rule of thumb I've seen elsewhere of using a trafo with VA rating 2.5 times the power usage of the amp) then it seems to me that the highest bias I could use would be around 1A. Say 800mA to leave a little headroom. If I've understand correctly (no guarantees of this!) then the amp will leave class A when the output current is twice the bias,
so that would be 1.6A, but that's an AC number, so RMS power is 1.6^2 x 8 / 2 = 10.2 W into a 8R load. (Sound right?)
This is more than I think I usually use while listening, but since this thread is at least in part about class AB, suppose I did use more. How do you estimate power into class AB? Obviously I would have rails much higher than needed, around 32V, say, after filtering and maybe a cap multiplier (haven't thought it through yet). If the rails were the limiting factor then you could swing 64V p-p, which would mean an AC current of 8A, and RMS power of 248W, which is way too high to be realistic and too much for the transformer anyhow. So there should be plenty of headroom on the rail voltage. There's going to be heat loss from the continuous bias current; 32VDC rails at 800mA means 64 x 0.8 = 51.2 W to be dissipated when no music playing, and I guess close to that when there is music playing. So there's about 110W available from the transformer. Is that the limiting factor?
I'm sure there's enough power for my listening habits anyhow, and I may use the PSU I already have ready, but I'm curious. Thanks to anyone with the time to set a non-engineer straight.
And thanks to Patrick for the project and files.
Nigel
For this one I have a Pass-style PSU I built for a different M2 build a couple of years ago that I could use, but I'm wondering about doing something different. One of my weaknesses is e**y; in particular I find nice transformers at a terrific price irresistible. I'm like a magpie - see something pretty, gotta have it... "Oooohhh..... shiny!!!" So I recently picked up a pair of Avel Lindberg 160VA transformers (the old ones from the UK) with 25V-0-25V secondaries. AT less than $9 each and modest shipping I couldn't resist. So they're just a little small, but much higher voltage than the usual transformers you'd choose for Pass-style monoblocks. But this discussion about using the M2OPS in Class AB got me thinking...
Suppose I were to build M2OPS monoblocks with these trafos. After the recitifiers I'll have +35V - 0 - 35V before smoothing, so if I estimate I have about 70W to burn (using the rule of thumb I've seen elsewhere of using a trafo with VA rating 2.5 times the power usage of the amp) then it seems to me that the highest bias I could use would be around 1A. Say 800mA to leave a little headroom. If I've understand correctly (no guarantees of this!) then the amp will leave class A when the output current is twice the bias,
so that would be 1.6A, but that's an AC number, so RMS power is 1.6^2 x 8 / 2 = 10.2 W into a 8R load. (Sound right?)
This is more than I think I usually use while listening, but since this thread is at least in part about class AB, suppose I did use more. How do you estimate power into class AB? Obviously I would have rails much higher than needed, around 32V, say, after filtering and maybe a cap multiplier (haven't thought it through yet). If the rails were the limiting factor then you could swing 64V p-p, which would mean an AC current of 8A, and RMS power of 248W, which is way too high to be realistic and too much for the transformer anyhow. So there should be plenty of headroom on the rail voltage. There's going to be heat loss from the continuous bias current; 32VDC rails at 800mA means 64 x 0.8 = 51.2 W to be dissipated when no music playing, and I guess close to that when there is music playing. So there's about 110W available from the transformer. Is that the limiting factor?
I'm sure there's enough power for my listening habits anyhow, and I may use the PSU I already have ready, but I'm curious. Thanks to anyone with the time to set a non-engineer straight.
And thanks to Patrick for the project and files.
Nigel
The M2OPS is nothing more than just the OPS taken from the original M2, plus an opamp up front.
So for other details such as power supply etc., it is best to refer to the original M2 thread :
https://www.diyaudio.com/community/threads/official-m2-schematic.281520/
Best,
Patrick
So for other details such as power supply etc., it is best to refer to the original M2 thread :
https://www.diyaudio.com/community/threads/official-m2-schematic.281520/
Best,
Patrick
- Home
- Amplifiers
- Pass Labs
- The M2 Output Stage in Class A/B, and maybe a Power WHAMMY?