Showing 1 of 264 conversations about:
View Full Discussion
I was thinking of building a headphone amp based on a circuit like the one on which the Bravo V2 is based. Being kind of lazy, I just bought this version of the Bravo V2, instead. I'm comparing mine to an original version Objective 2 amp, which I built up from a PCB I got directly from NwAvGuy.
So, what do I think of this Bravo V2?
1) Its output impedance is too high. It loads down trying to drive my Fostex T50RP headphones with 60 ohm impedance. Driving those 'phones, the Bravo gets 'gnashy' sounding. Harsh highs, rubbery bass. It sounds kind of 'electronic.' However, driving my Sennheiser/Massdrop HD6XX 300 ohm headphones, the sound from the amp becomes acceptable. It's still a bit more harsh and electronic sounding than if I use the O2 instead. But at least it sounds on par with a typical PC laptop's headphone output when driving higher impedance headphones. That's a pathetically low bar to clear...
2) I'm not having any noise or channel imbalance problems. Mine seems good.
3) I expected the Bravo to sound mushy and 'tube-y.' Instead it sounds 'harsh,' even when driving high impedance headphones. I haven't tried any different 12AU7 tubes in it, which I will do. But unless the supplied Chinese-made 12AU7 is an utter piece of crap, I don't expect a different version 12AU7 will fundamentally alter the annoying sonic character of this amp.
I figure the amp is worth the cost of its constituent parts. At some point I'll try some of the mods people have suggested. But if you're shopping for a 'keeper' of a headphone amp, ignore this.
I'm still happy with my O2, so if an Objective 2 comes up as a drop again, I'd recommend that. This Bravo V2 doesn't sound anywhere near as good as the O2. Not even close. Sorry.
I tried a few different tubes and the best ones I came across was this one in my opinion https://www.amazon.com/Electro-Harmonix-12AU7-EH-Vacuum-Tube/dp/B000UMDYYW
it really fixed a lot of the issues I was having with it. I upgraded the caps on my V2 and the highs sounded clearer and it lost whatever rubbery bass it had (even if I put the crappy chinese tube back in). After changing the tube out I've used it for my Grados 225es, K7XX, and my friend's HD 650s and it sounds great on all of them (Admittedly much better on the HD 650s). I'd highly suggest switching out the tube or modifying it though. I've sunk about $150 into my tube and it sounds better than any of my other amps at this point.
Can you please elaborate on the caps changeing?
Add some pictures, where and which caps you bought..
Maybe a small brief description on how to?
In case anyone's looking to mod this V2:
- I don't understand why people spend dozens of dollars (or Euros) on any ECC82 or 12AU7 to 'tube roll.' It's a waste. The provided Chinese tube is adequate. If you want to improve the sound slightly, try swapping in any American 12AU7 or similar, or any European ECC82. You won't notice much sonic difference, but you may feel better--especially if you can find a pretty 'Bugle Boy' ECC82 or a nice Mullard. Those are pretty. (I have a few 12AU7s lying around, but had to reject two of them because they're noisy. I did find a good GE 12AU7A, and that's what I'm using.)
- The capacitors in the amp look OK. The ones in my amp look like they were made by Rubycon and Nippon Chemicon. If those are really made by those companies, then they're decent quality parts. They are 85degC rated parts. Upgrading to 105degC rated parts could be a little helpful. Also, the installed caps are 25V-rated parts. Upgrading to 35V-rated parts might make for a tiny improvement. Changing those to higher quality parts might make a tiny difference, but I would not expect any major difference. Probably nothing audible in the end.
- The one big attainable difference would be to replace the IRF630 output MOSFETs (source followers) with IRF510, IRL510, IRF530, or IRL530. Why? Because the input capacitance of the IRF630 is a whopping 800pF (picoFarads), while the IRF510 (and IRL510) input capacitance is MUCH lower at 180pF. High input capacitance in the MOSFET is likely to cause an audible high frequency roll-off when driven by a 12AU7/ECC82 run with only 16V on its plate and very low plate (anode) current. That means the 12AU7 will be running with high internal resistance (rp), which would appear to be in series with the output MOSFET gate (input). That series resistance (R) with parallel capacitance (C) will make a low pass filter (LPF). If the 12AU7 rp is 15k ohms, and the input capacitance of the MOSFET is 200pF, the -3dB down point of the resulting LPF will be about 50kHz. That means its -1dB down point would be within the audible frequency range. In other words, it will have slightly muffled highs. Maybe worse than slightly muffled, depending on the quality of the IRF630 used.
Now if we replace that IRF630 with a (known to be good-quality) IRF510, the input capacitance should be reduced to somewhere around 50pF. R = 15k ohms and C = 50pF, so f3 (-3dB down point) = 212kHz. The -1dB point will now be above 20kHz. That's a BIG difference in the highs!
One thing I've seen is that people say the IRF510 or IRL510 will run much hotter than the original IRF630. The current to that MOSFET is set by the LM317 voltage regulator IC, which is being used as a current sink (current regulator). The only schematic I could find for this amp (http://bilder.hifi-forum.de/max/415727/bravo-v2-schematic_691080.png) shows an IRF510 used as the output source follower, with the LM317 set to 167mA current. If there's 15V across the MOSFET, with 167mA (0.167A) drain-source current, that means the MOSFET would be dissipating 2.5W of heat. The supplied heatsink should be able to deal with that. It could be that the people who are replacing the IRF630 with an IRL510 are not adjusting the value of the current setting resistor for the LM317 (R7 7R5 in the schematic linked above). Perhaps it's the LM317 that runs much hotter after IRF510 MOSFETs are installed? Perhaps R7 can be increased in value (to 10 ohms?) so that there's less current being drawn by the IRF530, and less heat needs to be dissipated by the heatsink on the LM317.
I'll try to replace the IRF630 MOSFETs with a couple of IRF510 I have in a drawer somewhere, and report back.
If I can replace the 1000uF 25V caps with better ones that will fit, I'll do that too. That's the output (DC-blocking) cap from the MOSFET source (C4 in the linked schematic), which is directly in series with the signal to the headphones. A better part there should make an audible difference, but might not. The only better caps I have are Nichicon MUSE, which are bigger than the ones supplied, and might not fit physically. We'll have to see...
When you upgraded the capacitors, what parts did you use? What manufacturer? Did you change the capacitance values? Did you change voltage rating? What temperature rating?
Electrolytic capacitors are pretty generic, so unless the originally installed parts are terrible-quality fakes (which is possible), I don't see why replacing one set of electrolytic caps with another pretty much equal-quality set would make any audible difference. Perhaps if you used really good quality, low-ESR, high pulse current-rated caps...
I ran a computer simulation of this amp's circuit, which told me that the amp will generate lots of high-order harmonics (distortion that lends an 'edgy' or 'brittle' sound to the music). It's interesting that the simulation predicts this, because when I first got this amp, I wrote:
3) I expected the Bravo to sound mushy and 'tube-y.' Instead it sounds 'harsh,' even when driving high impedance headphones.
So... It seems the actual sound of the amp agrees with what the circuit simulator predicted. Thinking that the simulator might help me improve the circuit, I tried some things. Nothing helped. I'm sorry, this circuit appears to be beyond help. It's not worth the effort to modify it. I'm not sure what this thing is good for, other than looking kind of cool on my desk at work . Unfortunately, it doesn't sound any better than the headphone jack on my work computer.
Oh well. Moving on now...
Hello all, first of all thank you for all the info about this amp and all the mods. I´ve made all the mods (changing the mosfets, changing the capacitors for high grade audio nichicons and other mods on this website rockgrotto.proboards.com/thread/11555/bravo-v2-modifications-2017) besides power supply filtering and some of them that are more complicated. The last one I did was the bias adjustment mod, replacing the 3k to 1k resistance and adding two 5k pots. But now I'm in a problem.
Is it normal that the tube filaments inside are like orange and really hot? I've left 13.5 V in the Bias adjustment.
The IRL530N is like super hot even with another aluminum dissipator. The LM317AT is like medium hot.
The major problem is with crosstalk and distortion. I´ve made the mod of crosstalk and with the right channel muted the music is perfect. However with the left channel muted I ear alot of distortion and leaking in the left and distortion in the right channel... I´ve already changed cables and tried to search online but I can't find a way to make it go off. I'm gonna attach some pictures for you to see what you think is happening:
Sorry, I no longer have the Bravo V2 amp I bought way back when. It's gone.
If you want help troubleshooting, I'd recommend drawing a schematic that shows the circuit with your mods, and showing the voltages you measure at each node.
The 12AU7 heater should be as close to 12.6V as you can get. It's OK if it's a bit low, like 12V, but I wouldn't run the heater higher than 13V, as that would be likely to shorten the life of the tube.
As I wrote in my post, I found that no amount of modifications would be likely to make this design into a truly 'hi-fi' amp (in my opinion).
Good luck with this.