Hey Guys I have a question myself... I am playing with the CRCRC filter resistors in order to 1) increase the filter's effectiveness even more and 2) to lower the temp of the regulators (next step will be increasing the bias current of the output stage a bit). The transformer is the 22V version.
With the stock 5,1 Ohm resistors, I get a voltage of 31V after the filter.
With 20 Ohm resistors, I get a voltage of 27V after the filter.
So I'm pretty happy, and the reg run about 5°C cooler.
For the moment I did the job only on one of the regs (easier to compare). I noticed that the LED of the regulator fed by the 20 Ohm resistors (27V input voltage) runs significantly brighter than the other one (31V input voltage). The output voltage is OK on both regs (+16,7V and -16,75V).
I don't get it why the led runs brighter now... Does anyone have an idea??
With the stock 5,1 Ohm resistors, I get a voltage of 31V after the filter.
With 20 Ohm resistors, I get a voltage of 27V after the filter.
So I'm pretty happy, and the reg run about 5°C cooler.
For the moment I did the job only on one of the regs (easier to compare). I noticed that the LED of the regulator fed by the 20 Ohm resistors (27V input voltage) runs significantly brighter than the other one (31V input voltage). The output voltage is OK on both regs (+16,7V and -16,75V).
I don't get it why the led runs brighter now... Does anyone have an idea??
Thank you Denis! Does the voltage matter?
What do you think of this one?
https://www.mouser.com/ProductDetail/KEMET/RSBEC0100ZA00M?qs=Jv4FAWB%2B0HaC6N7ijMa15Q==
it's basically the same, just do what's easier for you: on one of the AC terminals of the bridge, somewhere on the case (make sure you have good contact if the case is painted or anodised, though) or on the earth terminal of the IEC connector. The current from the shield to earth will be in the picoAmp or nanoAmp range, so it really doesn't make any difference. The only thing to take care about is to NOT connect shielding to signal ground, but to earth.
@da9520 polystyrene caps are great. They are easily available in 5% range and they are very stable. I use them whenever I can. Also WIMA MKS is a solid choice.
My build. Bottom yellow cable from power inlet into chassis earth. Left yellow cable to input cable shield. Black cable to input signal ground.
Works beautifully - I can turn the volume all the way up with zero noise. Impressive.
My build is done. Layered birch plywood, steel bottom and top plates. Worked first time - sounds amazing. Drives my Audeze LCD-X very well. Huge upgrade from portable DAC/Amp I used before. Changed the basic LM833 op-amp to a OPA2134PA - big difference, more timbre and sound stage. Very happy.
Very nice!! I didn't try the LM833, but other opamps, and I also settled on the OPA2134PA - to me (subjective) it is also the one that sounds best with this opamp, for exactly the same reasons you are pointing out. But other opinions may differ, of course!
BTW, you could cover the inside of your wooden enclosure with tin foil (well, aluminium foil) for shielding. it should work beautifully with spray-on glue. (don't forget to connect the poil to earth.)
For those interested, I played a bit with different CRCRC resistors. I ended up with the following setup which will be definitive:
- I replaced the 4x 5,1Ohm resistors by 2x 22Ohm 3W in parallel for EACH. So you need 8x 22Ohm 3W resistors for the amp. On each position, the first resistor is located where the 5,1 resistor was before (tin leads through PCB holes), and the second one is soldered on the bottom side of the PCB (make sure you have enough clearance). So: 4 resistors on top of the PCB, and 4 resistors on the bottom side. don't cut of the tin leads of the resistors that go through the PCB holes immediately, use them to solder the bottom-side resistors. Once they are all soldered, you can cut the excess tin leads.
This setup drops another volt or so over the CRCRC filter, and the regulators now see about +/-29V on their input instead of +/-31V. The regs now run significantly cooler (50°C instead of 55-57°C), and the CRCRC is even more effective than it was before. So there's no downside from my point of view.
I use the standard 2x22V transformer (which stays at exactly the same temp as before - around 30°C), and the standard bias on the output stage (10 Ohm resistors for R16, R22, R29, R32)). Due to the reduced reg temp, you could plan a bit with higher biases on the output stage if you want. I don't need it and will keep it "stock".
I also swapped the caps C12, C17, C22, C25 from Nichicon UKZ fine gold and Silmic II (I have built 1 amp with each kind, it's easier for A/B comparisons) to bipolarized Nichicon Muse UES (the bright green ones). I don't see any improvement as the amp was already dead quiet before the swap, but now I sleep better at night, because Wayne said it would reduce the noise level by another 3dB. So, if you are building the amp and need to order some parts, go for the Muse UES, but if you already built it with other caps, I wouldn't waste time and energy.
I'm more than happy with the amp now and I now call the project done. Now it's time to enjoy music with the Whammy and move on to building my A2 monoblocs that are waiting for months...
Again, by deepest gratitude goes to you, Wayne, for making this astonishing amp available to the public, to Nelson for supporting DIY audio, and to the whole diyAudio community for the hints, tweaks and help!
Cheers,
Denis
- I replaced the 4x 5,1Ohm resistors by 2x 22Ohm 3W in parallel for EACH. So you need 8x 22Ohm 3W resistors for the amp. On each position, the first resistor is located where the 5,1 resistor was before (tin leads through PCB holes), and the second one is soldered on the bottom side of the PCB (make sure you have enough clearance). So: 4 resistors on top of the PCB, and 4 resistors on the bottom side. don't cut of the tin leads of the resistors that go through the PCB holes immediately, use them to solder the bottom-side resistors. Once they are all soldered, you can cut the excess tin leads.
This setup drops another volt or so over the CRCRC filter, and the regulators now see about +/-29V on their input instead of +/-31V. The regs now run significantly cooler (50°C instead of 55-57°C), and the CRCRC is even more effective than it was before. So there's no downside from my point of view.
I use the standard 2x22V transformer (which stays at exactly the same temp as before - around 30°C), and the standard bias on the output stage (10 Ohm resistors for R16, R22, R29, R32)). Due to the reduced reg temp, you could plan a bit with higher biases on the output stage if you want. I don't need it and will keep it "stock".
I also swapped the caps C12, C17, C22, C25 from Nichicon UKZ fine gold and Silmic II (I have built 1 amp with each kind, it's easier for A/B comparisons) to bipolarized Nichicon Muse UES (the bright green ones). I don't see any improvement as the amp was already dead quiet before the swap, but now I sleep better at night, because Wayne said it would reduce the noise level by another 3dB. So, if you are building the amp and need to order some parts, go for the Muse UES, but if you already built it with other caps, I wouldn't waste time and energy.
I'm more than happy with the amp now and I now call the project done. Now it's time to enjoy music with the Whammy and move on to building my A2 monoblocs that are waiting for months...
Again, by deepest gratitude goes to you, Wayne, for making this astonishing amp available to the public, to Nelson for supporting DIY audio, and to the whole diyAudio community for the hints, tweaks and help!
Cheers,
Denis
To Denis and others that helped with advice, and of course Wayne for the design - I am super happy with the result. It is exceeding my expectations by a large margin. I can't recall my Audeze LCD-X sounding this good with the NAIM Unity Headphone Amp I tested with when I bought it. The LCD-X always had great bass, but the mid-range was soft and it lacked sparkle. Now the bass is just cleaner and crisper, the mid-range more full and wholesome (especially voices) and the tops have more tingle. I might want a little more zing on the high-hats, but apart from that the sound is amazing. I will be tempted to listen to this rather than my main system (Conrad Johnson pre, Purify-based power-amp and Martin Logan electrostatics).
Super clean, no noise (even at full volume). No lack of power - I listen with the volume at around 9 to 11pm (with a zero at around 7pm) - I can't imagine running out of oomph anytime.
I might be tempted to play with some discrete op-amps in a few month from time - will ask for advice then again.
Enjoy the forum - thanks for letting me be part of it for while.
Cheers
Manus
Super clean, no noise (even at full volume). No lack of power - I listen with the volume at around 9 to 11pm (with a zero at around 7pm) - I can't imagine running out of oomph anytime.
I might be tempted to play with some discrete op-amps in a few month from time - will ask for advice then again.
Enjoy the forum - thanks for letting me be part of it for while.
Cheers
Manus
If you like the sound of the Whammy and like DIY, you should definitely check out the Pass Labs B1 buffer preamp (PCB and FETS still available from the official site). I built mine over 10 years ago and it's still the most transparent, natural-sounding preamp I have heard to date (and that includes some Mark Levinson and Vincent gear). It's as close to perfection as you can possibly get. The only downside is that it has no symmetric in/outs. But be warned, it's addictive.
hum...I run mine with the volume knob set mostly between 9 and 11am instead of pm... but with the same results. Maybe the 12 hrs difference is due to the different time zones we're in! 🤣
nope, I built the original one which I bought from the official passdiy.com store when it was released around 2008.
Denis - I am interested in the Pass DIY pre-amp. How do you power it? Is a 18V re-chargeable battery an option? Or what is the cleanest power supply that works well with this?
Also, is there a separate forum for this?
Also, is there a separate forum for this?
There are indeed some threads about the Pass Labs / First Watt B1 buffer preamp on this forum. The best "entry point" might be here:
https://www.diyaudio.com/community/threads/b1-buffer-preamp.124889/
Just to avoid getting off-topic on this Whammy thread, may I suggest you re-post your question there, I'll answer it right away.
https://www.diyaudio.com/community/threads/b1-buffer-preamp.124889/
Just to avoid getting off-topic on this Whammy thread, may I suggest you re-post your question there, I'll answer it right away.
My build has 2.7 ohm bias resistors, good for about 220 mA through the MOSFETs. With the next size up heatsink they run at about 50C on the TO-220 tab. Of course that pulls a lot of current through the PSU, so the CRCRC resistors were dropped to 1 ohm and the capacitors bumped to the largest that would fit, 9000 uF. The stock transformer will not be happy at that load; where it lived was replaced with a mezzanine board featuring gozinta from a larger transformer, a hefty bridge rectifier, a spot of snubbing, and places to plug in a couple of 300 mH chokes to make the PSU LCRCRC. PSUD sims say it should have 0.4 millivolt of ripple at full bias, confirmed with a scope trace.
C3 and C4 have 10 uF ceramics as bypasses on general principle, which should cause no problem with the parasitic inductance of the electrolytic capacitors (about which a longish post anon; turns out all kinds of things lurk in the ultrasonics).
nice! good thing you beefed up the transformer. With 200mA bias, the stock transformer would get to its limits for sure! you're still using the regulators from the original design? they could be the next bottleneck, rated at 1-1,5A. If you beefed up the transformer, I would increase the R values of the filter. the higher the value, the better the filtering (make sure you use adequate power resistors, though). In your case, I thing you need 5W or 7W ceramic resistors to be on the safe side. Otherwise, it's a pity to have such a nice sturdy PSU and not use the whole capabilities of the filtering circuit. when I was playing with the R values, I found out that increasing R further reduces the ripple current significantly (1'm running at 1,5mV ripple (peak to peak) on after the filter).
Nice job on the bypass caps, that can't hurt, for sure!
Did you compare your build with the stock amp in terms of audio performance?
EDIT: are you sure you are really running 200mA bias? Have you measured it? even on a larger heatsink, 50°C is extremely low temp for such a high bias. I suspect there a bottleneck somewhere else (my 1st guess would be the regs, as stated earlier) that would limit the bias current well below 200mA
Nice job on the bypass caps, that can't hurt, for sure!
Did you compare your build with the stock amp in terms of audio performance?
EDIT: are you sure you are really running 200mA bias? Have you measured it? even on a larger heatsink, 50°C is extremely low temp for such a high bias. I suspect there a bottleneck somewhere else (my 1st guess would be the regs, as stated earlier) that would limit the bias current well below 200mA