Bravo v2 - 2017 | DIY Audio Heaven

leauk
valued member

Posts: 17

Bravo v2 - 2017 Jan 30, 2018 13:44:17 GMT

Quote

Post by leauk on Jan 30, 2018 13:44:17 GMT

Hi all. Have been recommended this site from head-fi.org where this post also resides.

Firstly I just wanted to thank the many original v2 modding enthusiasts, especially solderdude for sharing so much useful information and for their detailed and dedicated input to the Bravo V2 mods. I've just got back into audio and being an ex EE was attracted instantly to the Bravo v2 and mods. Mods I see have been talked to death and I've read many many pages, with almost information overload and have a mod plan, but still a couple of questions are outstanding. I'm using HD 580's 300R impedance.

I've also included part links as this was one thing I failed to locate through forums and I spent HOURS searching (bear in mind I'm not familiar with Hi-Fi part sellers so lots of searching, oh and I'm UK based) - so I hope this helps others too.

Plan (based on solderdude's recommendation and mine beginning 0.x).

Questions:

1. After completing the mods listed below, is it still worth changing the tube? I'm thinking:
GE 5 Star NOS 5814A - any thoughts appreciated.

2. Bias adjust - ref: i157.photobucket.com/albums/t69/pinkfloyd2007_bucket/bias.jpg

My v2 version doesn't have the trimmers, seems to have 25k resisters instead (25k is what the pcb states, however they are actually 24k).

So where to fit the presets, between which pins, and what are the 24k resistors for?

13.5v for IRF510s [IMG] measured at L220 and R220 photobucket.com/gallery/user/pinkfloyd2007_bucket/media/bWVkaWFJZDozMTgwMTcyNQ==/?ref=

3. Date codes for GE 5 Star tubes? I've seen a few on eBay with the code MZ 188-5 Green and grey print. Going on this information:

While I worked at Cain, we had a list of the two letter codes, and we were at the middle of the alphabet for the two years I worked there. G, H and K were the common second letter - they were the manufacture years 81, 82 and 83. The first letters were a simple A-M that stood for the month of manufacture.

Also, the months were labelled different for different years, even vs odd perhaps. A-M and N-Z or thereabouts were the two month groups, always ascending. So, AL or NL was a January, 1983 tube, depending on the month code group for 1983.

www.videokarma.org/showthread.php?t=134068

So far I understand the 188-5 means GE and Owensboro Kentucky was the plant of manufacturer. MZ is possible Month/Year ..which would be August 1983? Is this correct?

UPdate - now if GE used the same as its appliance division:
products.geappliances.com/appliance/gea-support-search-content?contentId=16195

July 1988 or 1976 - I wonder how we tell when year codes roll?

So much conflicting information - thanks to anyone who comments.

The most sound changing in order of magnitude:

0.1: I've added a valve change to my list based upon the tube rolling thread:
www.head-fi.org/threads/bravo-v2-review-tube-rolling-mods-coming-soon.671688/

GE 5 Star NOS 5814A - any thoughts appreciated.

1: IRL510 (highs extension)

cpc.farnell.com/vishay/irl510pbf/mosfet-n-logic-to-220/dp/SC11134?ost=IRL510PBF&scope=partnumberlookahead&exaMfpn=true&searchref=searchlookahead&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fw%2Fsearch

2: crosstalk cut

3: bias voltage setting (depends on tube)
- add 5k and 1k in series
cpc.farnell.com/multicomp/mf50-1k/resistor-0-5w-1-1k/dp/RE05031?ost=RE05031&scope=partnumberlookahead&exaMfpn=true&searchref=searchlookahead&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fw%2Fsearch

cpc.farnell.com/bourns/3309p-1-502/trimmer-5k-9mm/dp/RE05264?ost=RE05264&scope=partnumberlookahead&exaMfpn=true&searchref=searchlookahead&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fw%2Fsearch

4: heater mod (distortion reduction)
- need to research this more, is it supply heater tubes with analogue based PSU?

5: output resistance change. - change to 27R to provide approx 10:1 output balanced drive (300R:27R)
cpc.farnell.com/multicomp/mf50-27r/resistor-0-5w-1-27r/dp/RE05010?ost=re05010&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fsearch

6: output caps
- change to NFG-890 - 2200uF 35V FG Nichicon type, UFG1V222MHM (Fine Gold)
www.hificollective.co.uk/catalog/nfg890-2200uf-nichicon-type-ufg1v222mhm-p-5323.html

7: input caps
- not really sure about these, both due to construction and worthiness (listed last on the priority list)
cpc.farnell.com/wima/mks2b042201f00kssd/capacitor-2-2uf-50v/dp/CA05616 (corrected)

Non-Sound:

1: LM317 to LM317AT - Three-Terminal Adjustable Regulator - LM317AT
cpc.farnell.com/texas-instruments/lm317at/v-reg-adj-1-2-37v-to-220-3-317/dp/SC08257?ost=SC08257&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fsearch

2: Replacement heatsink for LM317AT - Heat Sink, TO220/218, 12.9°C/W - 6391BG
cpc.farnell.com/aavid-thermalloy/6391bg/heat-sink-to-220-218-12-9-c-w/dp/SC10867?ost=6391BG&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fsearch

Thermal paste: Heat Sink Compound 25g Tube - 200001000 25GM
cpc.farnell.com/servisol/200001000-25gm/heat-sink-compound-25g-tube/dp/SAHEATSINKCOMP?ost=200001000+25GM&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fsearch

3: Power supply cap (exceptionally low ESR) - Panasonic Electrolytic Capacitor, 4700 µF, 35 V, AM Series, ± 20%, Radial Leaded, 18 mm - ECA1VAM472X
cpc.farnell.com/panasonic-electronic-components/eca1vam472x/capacitor-4700uf-35v/dp/CA05753?ost=CA05753&scope=partnumberlookahead&exaMfpn=true&searchref=searchlookahead&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fw%2Fsearch

Bought 2 so may put these in parallel - not sure if that's a real requirement as I would imagine the low ESR and voltage lift is the crux of the mod?

4. Switch - really struggling to find a 8mm width >1.5A SPDT with toggle left to right, ended up with a 12.5mm DPDT (gold contacts, gives 10A ish if used in parallel) so may have to offset forward slightly to avoid the input cap. Panel Mount Toggle Switch, DPDT, On-On, 5A, Gold Contacts - 1MD1T1B1M1RE.
cpc.farnell.com/multicomp/1md1t1b1m1re/switch-dpdt/dp/SW02854?ost=SW02854&scope=partnumberlookahead&exaMfpn=true&searchref=searchlookahead&ddkey=http%3Aen-CPC%2FCPC_United_Kingdom%2Fw%2Fsearch

Note: due to the heatsink and caps we need to raise the top perspex plate:

5. Nickel plated M3 spacer - Spacers M3 / 10mm, 25 Pack - R30-3001002
cpc.farnell.com/harwin/r30-3001002/spacers-m3-10mm/dp/PC00099

Lea

REF:
rockgrotto.proboards.com/thread/7751#ixzz55TQ9aBAC
rockgrotto.proboards.com/thread/5297?q=irl510
rockgrotto.proboards.com/thread/5297

www.head-fi.org/threads/bravo-v2-review-tube-rolling-mods-coming-soon.671688/

Schematic - although not 100% identical to the Bravo V2
cdn.head-fi.org/a/7194711.jpg

Last Edit: Feb 6, 2018 10:15:11 GMT by leauk

leauk valued member Posts: 17	Bravo v2 - 2017 Jan 30, 2018 15:02:22 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by leauk on Jan 30, 2018 15:02:22 GMT Possibly an incorrect group, feel free to move this is to an appropriate one.
	Last Edit: Jan 30, 2018 17:07:47 GMT by leauk

solderdude
Administrator

measureutternutter

Posts: 4,886

Bravo v2 - 2017 Jan 30, 2018 18:09:27 GMT

Quote

Post by solderdude on Jan 30, 2018 18:09:27 GMT

Welcome to the forum.

The 24k resistor is a grid resistor.
It is between the grid and ground.
The input capacitor is connected between the grid and wiper of the volume pot.
It forms a high pass filter with the 24k resistor and is there to lower scractchy sounds when adjusting the volume and blocks DC.

The bias voltage pot is replacing the 3k0 resistor.
Can also be a 10k pot (with a 220 Ohm in series) for a wider adjustment range.

The heater mod reduces the distortion at higher listening levels.
It consists of feeding the heaters with a small DCDC converter (switch mode, not linear) but you will need much bigger heatsinks for the LM317 in that case. Even bigger than those of the IRL's.

The input caps need to be at least of the same value as the one currently in there.
Don't remember but think it is 1uF originally.
I recommend 2.2uF to 4.7uF film type (non polar) or put a bipolar electrolytic in there..
don't use a 1nF capacitor as there will only be some treble audible.

The output resistor can be changed.
Changing it from 47 to 27 Ohm doesn't do much.
The HD600 gets a little bass lift from a higher output resistance.

The 1:10 rule is found quite commonly on the web.
Nonsense story.
You can lower the output R to 2.2 Ohm or increase it for a bit more warmth/bass in the HD600
Would not increase it past 100 Ohm or so.
Experiment or.... put a small 100 Ohm trimpot in there and adjust it !

Can't help you with the tube.

The Bravo amps have a tendency to go belly up within a year or two.
With most parts replaced it is less likely.
The PCB is of poor quality and pads or traces breaking or coming loose from the PCB is not uncommon when working on it.

When the amp gets all kinds of weird behaviour check the 'clamping force' of the tube socket and bend the contacts a bit closer again.
Would not tube roll too much as the socket is not really suited for this (quality wise)

Last Edit: Jan 30, 2018 22:53:04 GMT by solderdude

use your EARS to enjoy MUSIC .... NOT as an analyzer

oldson
extremely active

Posts: 1,678

Bravo v2 - 2017 Jan 30, 2018 19:13:23 GMT

Quote

Post by oldson on Jan 30, 2018 19:13:23 GMT

Hi Leauk,
welcome,
hope you dont mind me chiming in?
i am not knowledgeable at all in this field and dont have the skill to play around too much on the diy front, but i have to ask....
have you not considered dumping the bravo and buying a kit from G1217?
i would have thought this would be right up your street, considering you know your way around circuits.
never seen anyone criticize their value or quality.
cheers

Headphones: JMA XTC open, DT1990,Hd650,Ah-d2000,DT770,MDR-z7m2,amiron home.
Dac: Chord Qutest, Topping E30
Rega P8/Ania-pro
Rega Aria phonostage
HeadAmps: Ember v1,SPL phonitor se, icon hp8
schiit lokius
Amp:Rega Elex-r,
Dynaudio emit M10 speakers

lobalwarming
quite active

Posts: 170

Bravo v2 - 2017 Jan 30, 2018 23:15:04 GMT

Quote

Post by lobalwarming on Jan 30, 2018 23:15:04 GMT

Agree with oldson. Building a G1217 amp will permit the rolling of a wider variety of tubes. While the 12AU7 is a decent family, it's the meat and potatoes of tube rolling. The 5814s add a little more spice - I enjoy the vintage RCA and Tung-sols 5814s. To a point.

But there's so many more delicious flavours to taste. And the G1217 amps (Solstice, especially) allow a more satisfying palette of tube rolling options. Once you've sampled exquisitely defined full-sized gourmet sonics from a low cost 12SN7/6SN7 it's tough to go back to mere 12AU7 burgers every night.

leauk
valued member

Posts: 17

Bravo v2 - 2017 Jan 31, 2018 13:44:12 GMT oldson likes this

Quote

Post by leauk on Jan 31, 2018 13:44:12 GMT

Hi oldson and lobalwarmin, many thanks for your reply. I had come across this beautiful tube headphone amp range (G1217) but the initial attraction to the Bravo v2 was:

I spotted the Bravo v2 on Amazon and its cost is great for newcomers - and speed of delivery ;-) You know what it's like!

I discovered after making that instant purchase click, the huge resource of information across forums.

Plenty of well documented mods, admittedly due to failings, but you learn so much from failings...I have educated myself within headphone audiophile amps and tubes well beyond that I would have having a plug-n-play device.

And I'm an ex EE, so modding and redesign is VERY close to my heart and well within capability and tool ownership - mind you had the board be comprised of <1206 SMD I would have been out!

I'm not really into tube rolling to be honest, just heading back into audio from a break, but am keen to learn.

And, hey it's my first audiophile project, learn cheaply and then move upwards.

Oh and my GF loves the look (and cost), anything in the lounge has to be 'approved' ;-)

Cheers
Lea

Last Edit: Jan 31, 2018 13:49:06 GMT by leauk

leauk
valued member

Posts: 17

Bravo v2 - 2017 Jan 31, 2018 15:02:37 GMT

Quote

Post by leauk on Jan 31, 2018 15:02:37 GMT

Thanks solderdude, sound advice indeed (forgive the pun).

The 24k resistor is a grid resistor.

It is between the grid and ground.

The input capacitor is connected between the grid and wiper of the volume pot.

It forms a high pass filter with the 24k resistor and is there to lower scractchy sounds when adjusting the volume and blocks DC.

Wish I had a schematic so I could review. I have little understanding of valves :-(

The bias voltage pot is replacing the 3k0 resistor.

Can also be a 10k pot (with a 220 Ohm in series) for a wider adjustment range.

Perfect, thanks.

The heater mod reduces the distortion at higher listening levels.

It consists of feeding the heaters with a small DCDC converter (switch mode, not linear) but you will need much bigger heatsinks for the LM317 in that case. Even bigger than those of the IRL's.

Not sure I'll bother (unless you mention otherwise) as I rarely listen at high levels.

The input caps need to be at least of the same value as the one currently in there.

Don't remember but think it is 1uF originally.

I recommend 2.2uF to 4.7uF film type (non polar) or put a bipolar electrolytic in there..

don't use a 1nF capacitor as there will only be some treble audible.

Yes, 1uF standard electrolitic. Film? I was planning silva mica. Cheers.

The output resistor can be changed.

Changing it from 47 to 27 Ohm doesn't do much.

The HD600 gets a little bass lift from a higher output resistance.

Actually this was one of your previous suggestions, but I thought at the time 47R to 27R is such a little change compared to 300R.

The 1:10 rule is found quite commonly on the web.

Nonsense story.

You can lower the output R to 2.2 Ohm or increase it for a bit more warmth/bass in the HD600

Would not increase it past 100 Ohm or so.

Experiment or.... put a small 100 Ohm trimpot in there and adjust it !

I may just increase it as the HD580s are a little lacking below 60Hz.

The Bravo amps have a tendency to go belly up within a year or two.

With most parts replaced it is less likely.

The PCB is of poor quality and pads or traces breaking or coming loose from the PCB is not uncommon when working on it.

Yes...the power switch and input cap voltage is atrocious design.

When the amp gets all kinds of weird behaviour check the 'clamping force' of the tube socket and bend the contacts a bit closer again.

Would not tube roll too much as the socket is not really suited f

Great tip, thanks. I've seen other threads reapalcing the socket too, especially for rolling (which I don't really intend to do).

Big thanks again
Lea

Mr Trev
very active

Posts: 386

Bravo v2 - 2017 Jan 31, 2018 16:44:01 GMT

Quote

Post by Mr Trev on Jan 31, 2018 16:44:01 GMT

Jan 30, 2018 23:15:04 GMT lobalwarming said:

Agree with oldson. Building a G1217 amp will permit the rolling of a wider variety of tubes. While the 12AU7 is a decent family, it's the meat and potatoes of tube rolling. The 5814s add a little more spice - I enjoy the vintage RCA and Tung-sols 5814s. To a point.

But there's so many more delicious flavours to taste. And the G1217 amps (Solstice, especially) allow a more satisfying palette of tube rolling options. Once you've sampled exquisitely defined full-sized gourmet sonics from a low cost 12SN7/6SN7 it's tough to go back to mere 12AU7 burgers every night.

Darn it Lobey, there you go on the SN7's again. As if their price isn't inflated enough, you keep trying to get more folks hooked on the things

I'll second Solderdude's remarks about the poor PCB and tube socket. I lifted some traces when modding my Little Bear variant. A couple of tubes changes is about all it took to loosen the contacts on my socket which led me to replace them - which led to the lifted PCB traces. A delicate soldering hand is pretty much necessary…

Last Edit: Jan 31, 2018 16:50:23 GMT by Mr Trev

solderdude Administrator measureutternutter Posts: 4,886	Bravo v2 - 2017 Jan 31, 2018 17:03:57 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by solderdude on Jan 31, 2018 17:03:57 GMT schematic: Silver Mica does not exist above 100nF The one you selected was 1nF, not 1uF The input caps must be film type. 2.2uF, 3.3uF or 4.7uF
	use your EARS to enjoy MUSIC .... NOT as an analyzer

leauk
valued member

Posts: 17

Bravo v2 - 2017 Jan 31, 2018 17:46:22 GMT

Quote

Post by leauk on Jan 31, 2018 17:46:22 GMT

Thanks for spotting I'd forgotten pf/nf/uf conversion - see it's been a while! I would blame it on fat typo fingers, but hey honesty and policy and all that!

This should do the trick ?
cpc.farnell.com/wima/mks2b042201f00kssd/capacitor-2-2uf-50v/dp/CA05616

Thanks for the schematic (assuming 12au7 tube is interchangeable), I spotted Ocean schematic on diyaudioheaven and it's remarkably similar bar the PSU decoupling cap.

I see the 24k/1u HPF now. You mention 2.2uF or above - to increase the cut-off and push the -3db point lower so at audio we're higher up the curve?

Cheers
Lea

solderdude
Administrator

measureutternutter

Posts: 4,886

Bravo v2 - 2017 Jan 31, 2018 18:05:16 GMT

Quote

Post by solderdude on Jan 31, 2018 18:05:16 GMT

Yep... 1uF = 6Hz high pass (-3dB) but the output caps also form a high pass and is why I recommend to increase both input and output caps in value.

2.2uF = 3Hz and 4.7uF = 1.5Hz -3dB points.

One can easily assume these low frequencies is moot but the -0.5dB is much higher so increasing the value is a good idea.

All these amps are very similar and equally flawed in design.
Still, thanks to our ears being not as good in detecting distortion as people think they are these amps are well sold and liked.

use your EARS to enjoy MUSIC .... NOT as an analyzer

leauk valued member Posts: 17	Bravo v2 - 2017 Feb 1, 2018 13:51:26 GMT Quote Select Post Deselect Post Link to Post Member Give Gift Back to Top Post by leauk on Feb 1, 2018 13:51:26 GMT One can easily assume these low frequencies is moot but the -0.5dB is much higher so increasing the value is a good idea. That'll be the first-order dilemma! sim.okawa-denshi.jp/en/CRtool.php

solderdude
Administrator

measureutternutter

Posts: 4,886

Bravo v2 - 2017 Feb 1, 2018 15:57:41 GMT

Quote

Post by solderdude on Feb 1, 2018 15:57:41 GMT

because there are 2 high pass filters the dilemma could be between first-order and second order depending on the headphone impedance.

frequency -3dB point is easily calculated.
1/(2 x 3.14 x R x C) where R is in Ohm and capacitance in Farad, freq. in Hertz

As they are high pass filters this would be the correct tool:
sim.okawa-denshi.jp/en/CRhikeisan.htm

Even though the frequency response may well extend to say 6Hz at - 3dB the - 0.5dB point is higher up, just below 20Hz.
Still, that would seem enough but the phase response is still 'off' till about 60Hz or so.
For this reason I always design stuff with extended bass response.
Over dimensioning does not hurt in this case.

For output capacitors this is even more the case as the output impedance rises at lower frequencies as well.
2200uF seems overkill, certainly for 300 Ohm headphones but for 16 Ohm or 12 Ohm headphones the 2200uF is already a bit small for my taste.

Last Edit: Feb 1, 2018 16:10:51 GMT by solderdude

use your EARS to enjoy MUSIC .... NOT as an analyzer

leauk
valued member

Posts: 17

Bravo v2 - 2017 Feb 2, 2018 10:36:47 GMT

Quote

Post by leauk on Feb 2, 2018 10:36:47 GMT

Whoops, I posted the LPF link accidentally - thanks.

2 high pass, 1st at the input, 2nd the output cap?

Even though the frequency response may well extend to say 6Hz at - 3dB the - 0.5dB point is higher up, just below 20Hz.

Indeed, hence mentioning to push the audio response up the curve

Still, that would seem enough but the phase response is still 'off' till about 60Hz or so.

I was going to ask you about the phase....

Also, it's tricky calculating the -0.5db point (well for me) as we're on the knee outside the 6db per Octave linear quadrant - maybe you have a handy formula?

For this reason I always design stuff with extended bass response.

Over dimensioning does not hurt in this case.

Sounds wise :-)

For output capacitors this is even more the case as the output impedance rises at lower frequencies as well.

2200uF seems overkill, certainly for 300 Ohm headphones but for 16 Ohm or 12 Ohm headphones the 2200uF is already a bit small for my taste.

So for 300R there's little point changing the output caps?

Thanks for all your advice
Lea

solderdude
Administrator

measureutternutter

Posts: 4,886

Bravo v2 - 2017 Feb 2, 2018 21:20:05 GMT

Quote

Post by solderdude on Feb 2, 2018 21:20:05 GMT

Feb 2, 2018 10:36:47 GMT leauk said:

2 high pass, 1st at the input, 2nd the output cap?

The input side has the input cap (1uF) and 24k as high-pass.
On the output side C out and the connected headphone + 47 Ohm // 1k is the second high pass filter.
This only is an issue when connecting low impedance headphones though.
Not for 300 Ohm headphones.

I was going to ask you about the phase....

Also, it's tricky calculating the -0.5db point (well for me) as we're on the knee outside the 6db per Octave linear quadrant - maybe you have a handy formula?

Say a factor 2.5 will get you in the ballpark figures.

For output capacitors this is even more the case as the output impedance rises at lower frequencies as well.

2200uF seems overkill, certainly for 300 Ohm headphones but for 16 Ohm or 12 Ohm headphones the 2200uF is already a bit small for my taste.

So for 300R there's little point changing the output caps?[/quote]

correct. the output high-pass is 0.5Hz in this case

use your EARS to enjoy MUSIC .... NOT as an analyzer