Bravo 12AU7 Mods and Appreciation thread

[Garage1217]

Great work!

[Deleted]

I would say make a board for a modded G2 (separate heater power supply with 12V-6V switch and output disable relay)
Also make the cathode pot a 5k one with multiturn or better yet leave out the cathode pot (and by this eleminating the need for a cathode electro) and make the anode CCS adjustable instead.

[Garage1217]

Thanks for the feedback Solderdude! I was actually logging on to message you for your input as I value it after reading your posts. I would be happy to share the double sided board layout I am working on with you to see what you have to say and get your input on the design. Also, I already spec'ed the bias pots at 10K 25 turn 1/2 watt bourns units, simply because of the way my fav tube runs, I needed more resistance as the 5K did not cut it. A 6V switch is a great idea as well.

Here are the mods I am implementing so far...
- Double sided board
- 25V 4700uF fake rubycon power supply cap changed to a 3qty of 2200uF 35V Nichicon KT series caps for a total of 6600uF..
- RCA jacks removed / only 1/8 jack input?? just a thought at this stage or maybe just have rca's instead? No need for both IMO as it clutters the board. Also re-route the input to the back of the board for a cleaner look
- Crosstalk mod built in / circuitry re-routed
- Input signal capacitors eliminated rather than removed and jumpered. Or maybe leave a provision for them and a jumper to eliminate?
- Blue LED turned WAY down brightness wise with a 10K resistor + 100uF 10V cap added for soft start
- 25 turn 10K 1/2w bourns bias pots for precision bias adjustment. Along with easy to access points to check and change bias with a standard multi-meter
- Output signal capacitors changed to Nichicon 2200uF KT series caps
- Larger heatsinks for both 530's and 317's / larger more open board design at 4x6"
- IRF630's changed to IRL530's or IRL510's for massive extension of the top end / rather than cutting falling off at 10K
- LT317's changed to Linear technologies LT317A's which are of better quality
- All resistors will be 1% Vishay / dale mil spec units
- Power switch maybe located on the top lid portion of a custom laser cut acrylic chassis (similar to the current design)but with a quality switch
- Improved symmetrical board layout / I would like to add an artistic touch to the look and layout without sacrificing functionality
- Debating on just having solder points for the volume potentiometer as I am not a fan of the stock alps unit... so one could choose what style they want. Same with the 1/4" output jack... The 1/4"stock unit is not bad and common so could leave it on the board. Still debating with myself on it.

Keep the feedback rolling, Want to make this the best unit I can.

[duheee]

Hello everyone. I am kind of new when it comes to electronic components, so i hope that someone can answer some questions that i have.

My questions are:

How do I cut the traces for the crosstalk improvement mod? Do I just use an X-acto knife, and then solder onto the already existing joint? Also, will any wire work for it? What is the motive for doing this mod; what type of improvements will i hear?

I noticed that the mosfets seem to be glued onto the heat-sinks. Does it take much effort to get the mosfets off? Also, it appears that the mosfets are soldered on the top as well as the bottom of the board. Is it necessary to solder on BOTH sides, or is just soldering on the bottom side acceptable?

Thanks,
Duheee

[Deleted]

Hi Duhee,

Welcome to the forum.

Yes, cut with a X-acto knife and solder wires from existing pads to other pads.
It will eliminate the crosstalk.
This you can make quite audible by only plugging in the L or R input channel and listening to the channel that is supposed to be silent.

The MOSFETS are not glued on the heatsinks.
The white stuff is cooling paste for better thermal contact.
reuse as much of the stuff as possible between the MOSFET and heatsink.
DON't get it on your clothes !

When soldering the MOSFET it should flow through the hole and be visible on both sides.

This is how I get them out:
unscrew the MOSFET from the heatsink and lift it off from the heatsink so you can grab the top with small pliers.
Apply LOTS of solder on the 3 legs so you get one solder blob that can be heated as a whole so all 3 legs have molten solder.
When the solder has melted GENTLY pull the MOSFET out.
Clear the holes with solder sucker or by heating it and blowing the holes open with compressed air.

[duheee]

Unfortunately, after attempting to desolder the biggest capacitor to replace it, I ran into trouble. My soldering iron was unable to heat the solder and make it molten. I tried for about an hour to heat the connections, but to no avail (leaving a bunch of scratches around the pad in the process). I finally got the cap off, but only after pulling it until the connection snapped, and now the connections are left in there. Is there something wrong with my soldering iron? It is 15W. Also, is it because of the lead free solder, or am I just using the wrong technique? Could somebody explain the process of desoldering this because the way that I was taught is not working.

[Deleted]

15 W is good for soldering single sided boards and 60/40 solder.
It is underpowered for desoldering.

You will need a (cheap) temperature controled station for this or 30W/40W iron with a small chisel tip.
Also you would need to add some 40/60 solder.
IF leadfree was used you need higher temperatures.
Especially if you plan to desolder the IRF's as the middle pin is coupled directly to the heatsink and a small iron simply wouldn't heat the joint enough to melt.

with a more suitable iron (don't need to buy an expensive one) heat the wire that is stuck (add some solder) and pull it out with pliers.

[duheee]

I successfully desoldered everything. I found a 25w iron, desoldering wick, and a can of air, and made good use of them . I soldered the mosfets in successfully. Now, it's time for the 5k trim pots and the caps .

Thanks for the help,
Duheee

[duheee]

I can't get the trim pots to go in properly; the legs are too small. Is there a secret to getting them in? I also accidentally snapped off one of the legs , but after a quick chat with mouser, they are sending me a new one free of charge. I guess i will have to wait a few more days to complete the project...

[PinkFloyd]

Jun 20, 2011 23:15:59 GMT duheee said:

I can't get the trim pots to go in properly; the legs are too small. Is there a secret to getting them in? I also accidentally snapped off one of the legs , but after a quick chat with mouser, they are sending me a new one free of charge. I guess i will have to wait a few more days to complete the project...

It would appear to me that you haven't got a clue what you are doing and I would suggest you stop NOW before you totally fekk the amp up.

Sit back, practice soldering / desoldering on a scrap board and learn a bit more about the art of soldering....

Seriously mate, I am not going to encourage you to go any further and would recommend you DOWN TOOLS until you know what you are doing

[PinkFloyd]

I tried for about an hour to heat the connections

SERIOUSLY mate...... stand back and ask yourself "am I wise?".... Stay well away from the amp and seek professional advice.

[duheee]

I took a class on soldering. The only downside about this is that i was not tough how to desolder (maybe not the best class); although I learned the basics on tangent tutorials. How are you supposed to learn if you never attempt something? I have done my share of practicing on a spare board, but have never dealt with some of these components before. I don't have spare 5k trim pots laying around. Also, how am i supposed to know if i am doing something right if i don't ask someone who has done it before? Why do you think I am asking all of these questions? Exactly because I don't want to "totally fekk the amp up."

SERIOUSLY mate.........It seems to me that you need to stand back from the keyboard and ask yourself "Why is duheee asking all of these questions?" ( because he is trying to learn, if you didn't figure that one out)

I am not trying to be mean, but you should consider what you are typing before you type it.

Duheee

[Deleted]

> I can't get the trim pots to go in properly <

Solder some wires to the pins.
Then insert the pot and, very quickly, solder the elongated wires to the board.
Check to see if the extra wires did not unsolder themselves from the pot when you soldered it onto the board.
I use leads clipped of parts I already fitted for this.
The longer these wires are the less likely the pot is desoldered by accident when soldering the wires to the PCB.
The long legs help disspate away the heat.

SET the pots to the middle position when firing up the amp and wait 30 seconds before adjusting the pots.

b.t.w. Mike (Pink Floyd) has seen many amps ruined by 'novices' soldering actions and is merely trying to posuede you to do some 'practice soldering/desoldering' first on expendable stuff to prevent ruined PCB's you don't want ruined.

The Chinese Indeed/Bravo etc PCB's are not very high quality IMO and traces pads come of quite easily.

[tomco]

I picked up an "XY 6N11" amp on ebay a while back. I'm interested in doing some mods but the PCB layout doesnt seem to match the Bravo/Indeed amps despite being cosmetically similar.

Does anyone know what this amp is a clone of?

[Deleted]

It is a 12AU7 bravo alike thingie with some improvements.

a: cooling is better (heatsinks on the outside and the LM317 heatsink is bigger, though not as big as the surise)
b: on/off switch at the front instead of back
c: 3.5 mm AND 6.3 mm output socket
d: the bias voltage control pots are there (as suggested in this thread and almost as applied in the Sunrise project, the sunrise one is a bit better adjustable)
e: it appears to have bias test points (like in the sunrise project)

it does, however, still have the crosstalk problem (sunrise hasn't).

What MOSFET's are used ?
What tweaks are you looking for (cathode bypass caps are not needed)
can't make it out but the output resistors appear to be 10 Ohm or 100 Ohm (the middle ones between the 3.5 socket and volpot) measure them Ohmic with the amp off and disconnected.
166 mA class-A current
the output caps should have been mounted away from the tube, heated caps dry out quicker.

[tomco]

It uses a pair of IRF630N's and LM317T's.
The input cap is an Elna 25v 1800uf
The output caps are Rubycon 25v 1000uf.

I haven't got a multimeter to hand, but reading the codes the output resistors are from left to right: 110, 100, 100, 110.

I want to try the mosfet replacement, crosstalk mod and possibly change the caps.

Could I mount new output caps sideways to move them away from the tube?
Maybe one over the 3.5 socket and the other over the output resistors. Could get a bit messy.

[Deleted]

I suspect the outer ones will be 1k (brown, black, black, brown, brown) and the output resistors 100 Ohm (brown, black, black, black, brown)
difficult to tell with certainty from the pics.

crosstalk mod, but still using the heaters, as in the the first page of this 12AU7 thread, cannot be done as this amp only takes 6V tubes and not the 12V types.
It is more like the original bravo/indeed/miridiy thingies in that aspect.
It can only be done by adding the extra power suplly (as described in the G2 article) and cutting some traces.

The IRF630 I would replace with IRL510 (or IRL530 if you want to retain a bit of it's current character)
adjust the bias voltage to 13.5V with IRL fitted, with IRF 15.5V
The input cap I would replace with a 35V type (around similar value but higher voltage rating) mind the diameter and pin distance (footprint).
The output caps I would replace for 2200/25V type, in this case you can lower the output resistors a bit too (do not go lower than 47 Ohm)

The 2200/25 caps will be bigger and/or have a bigger diameter.
You can mount them away a bit from the tube on longer legs.. would secure them somewhere (tie-wraps or hotmeld glue?) in that case.

[tomco]

Thanks for the info, I'll see if I can place an order for the components during the week.

Just a bit of clarification needed here:

The IRF630 I would replace with IRL510 (or IRL530 if you want to retain a bit of it's current character)
adjust the bias voltage to 13.5V with IRL fitted, with IRF 15.5V

So with the current IRF630 I should adjust the bias to 15.5V? I thought it was 16.5V?

Also, is there anything I can do about "noise"? I'm not really sure where the noise is coming from but I'm getting a rythmic clicking noise and buzzing. The amp seems to be quieter in some places and noisier in others.

Am I right in thinking this is interference to the tube from a wireless device?

[bizzie]

I had a rythmic ticking noise when debugging a guitar amp that I had built and and noticed one day that it only happened when on the bench with my left hand near it, turned out to be my digital watch

[tomco]

Fantastic, my amp just killed my faithful zen vision:m and the 30gb of music contained therein
I'm thinking about some sort of Cowon as a replacement (i9/s9).

Anyway, I noticed the other day when I was adjusting the bias that there appeared to be around 20V's running across the exposed heat sinks! I'm not the most experienced at this, but I'm thinking this is not normal!

Or course I promptly forgot about it.

Today I was testing my new DT150's (courtesy of PinkFloyd) and set up the amp with the closest interconnects to hand (bad idea). I used a 3x RCA to 3.5mm from some video device or other and left the spare RCA loose.

Yep, the ring on the spare RCA met the heatsink - huge spark - dead mp3 player.

Any idea how the hell the heatsinks have 20V's running across them.
I'm not going anywhere near the amp until I get this sorted (I'm afraid its going to kill me, or something expensive!)

[Deleted]

If you use isolation sets on the MOSFET's there is no 24 Volts on the cooling fins anymore.

From Solderdude (reply 53)


e.g.
www.jaycar.com.au/productView.asp?ID=HP1176&keywords=heatsink&form=KEYWORD

[tomco]

Crap!

Guess I shouldn't have skimmed the thread

[ohmer]

So, I managed to change caps and transistors. It actually works. Do you guys have a trick for removing 3 pin transistors? I just wrapped a wire through the screw hole and then bathed the pins in solder so it would flow out together and then cleaned up the excess.

The traces suck balls.

Sounds sooo much better. It seems to kick more on the low ends.

Now I'm looking to change the resistors with the trim pot for bias.

I have Bourns 3295 and according to the data sheet W103 on the top means 10k ohm.

So my questions are, will this cause problems for me if I venture into the pot min and max and It has 3 pins, do I connect the center pin and either of the outer pins.
I guess the middle pin is the moving wiper and either end is the stationary connect?
I see it in the data sheet but I want to make sure I do it clean.

I assume its safe to clip the left over?

Thanks

[maxloud]

Hi and a big thank you for all your contributions in this wonderful forum! I am completely new here as a registered member, but have been following some very informative discussions over the last months.

I also followed the detailed instructions to mod my Bravo Audio V2 and got some very nice results - until yesterday, that is.
I got a somewhat exotic version of the Bravo tube amp, because the bias adjustement pots are already present. So I exchanged the 630 mosfets and installed the 510s and I also exchanged the 25V 4700 µF cap against a more sensible 65V 6700 µF version. (I couldn't get the 35V version, but thought, the 65V would be as fine).

I also tried several tubes besides the Electro Harmonix, which was factory installed:
- an old Valvo, which I found quite boring
- a new TAD, which is very clear and crisp, but not so much different from the Electro Harmonix.

Installing the 510 Mosfets made a very big improvement, much bigger, than my somewhat lacklustre trials with the tubes.

But, after doing the mods, I had some audible noise, which I found quite unnerving. The Bravo also developped some serious distortion, when playing a bit louder.

I thought, there might be two sources for the noise and distortion: may be the 6700 µF Cap was too much and ofcourse the desoldering of the parts with the lead free solder was a distaster. (I am not a novice at soldering, but never have worked with lead free boards before). So some of the solder points are very poor. I bridged the poor soldering point yesterday with some wires, especially from the Mosfet contacts to the next best matching other soldering point on the board, which went fairly easy. I also replaced the 6700 µF cap with a 4700 µF type (I got that in 35V), which is also sizewise a much better fit.

After this last "improvements", the noise in the left channel was extreme - so much so, that it is impossible to use the amp anymore. I then measured the bias and found 23.5 volts, compared to 11.2 volts on the right channel. Unfortunately the 23.5 volts stay put and changing the left channel pot does not change the voltage at all. I veryfied, that both pots work the same and the right channel (which does not exhibit this excessive noise) can be adjusted bias-wise.

I traced my soldered wires and cannot find any short.

Is there a possibility, that the fixed overly high bias voltage is due to a tube failure? or have I killed the left channel Mosfet while soldering. There is also a transistor in each channel - could that be the culprit. I am somewhat out of my wits, as I have never worked on a tube thingy before. (Yes, you can heap blame on me, for being stupid, but there is nothing good, unless you do it...)

Thanks for any pointer into the right direction!

Max

EDIT - 2 days later: while writing the above epistle, it occured to me, that the 23+ volts are app. what the power supply would deliver. This made the search for a short or wrong connection much easier and today I sat down and found my mistake within 5 minutes. I bridged two solder points, which I shouldn't have used and such layed the PSU voltage right to the tube socket, bypassing the bias adhjustment. After removing that wrong connection, everything is fine. Also, my wired bridge obviously work and the amp now sounds clean and beautiful. I rund the TAD tube with 16.5 volts on both channaels currently. Perhaps I will try other voltages during the next day to find the sweet spot.

[pedantyc]

Hey Solderdude, thank you very much for your help and instructions on here. Especially for that pdf of the mods. Very detailed and helpful. I did have one question though. When I replaced my output caps on the Muse amp I am working on, I noticed that the original caps were 35v ones. I replaced them with 25v/1000uf. Should I leave it this way, or should I run out and grab some 35v caps?

Also, if cutting the traces on the bottom of the board and adding that wire, would I still need to relocate the lm317AT's? I wasn't keen on having to do that or adding the TSR 1-2465, but I wasn't sure if cutting the traces is what causes the large amount of heat or if it was the addition of the regulator.

Thanks again for the guides, never would have tried this without the help.