Bridge construction

[GrahamnDodder]

No, not drawbridges, or suspension models.

I am talking about the diode kind - specifically with the mur860's that I have ordered.

These, I believe, are the heavy-duty-looking type, shaped like the LM3875, but smaller with two pins.

I know how to wire the bridge electrically, but physically, how is the best way to construct the bridge?

Do I mount them any way special? Are there heat issues? Can I solder the appropriate pins together with the wires and just build it with no add'l structure?

If anyone has a photo of a way in which they did an 860 bridge and they were very pleased with it, please do post.

Thanks,

GND

 

[rwagter]

Here you are:

An externally hosted image should be here but it was not working when we last tested it.

schematic:

An externally hosted image should be here but it was not working when we last tested it.

it's made by Peter Daniel the original posting is here:

http://www.diyaudio.com/forums/showthread.php?s=&postid=61921#post61921

Ralph

 

[GrahamnDodder]

Wow. I dunno if I am up to Peter's level.

Is it really necessary to heatsink all of the diodes? (I am running 25/25/330VA)

Is the hole isolated? If not, is it A or K?

GnD

 

[Mad_K]

No need to heatsink them in a gainclone. (The pic above is from Peter's AlephX i believe).

Here's a few ideas:

http://diyaudio.com/forums/showthread.php?postid=185161#post185161

http://www.diyaudio.com/forums/showthread.php?postid=170244#post170244

😉

 

[mikelm]

No need to heatsink them in a gainclone.

Talking about heatsinking, what material are you using for heat tranfer isolation washers ? It looks like ceramic.

if it is where did you get it from ?

cheers

mike

 

[Nuuk]

GND - your question is answered (simply and fully I hope) on my GC FAQ page .

 

[carlosfm]

No need to heatsink them in a gainclone./B]

While testing them, I found out that they do get hot.
I wouldn't use them without a heatsink, even a small one may do.

 

[Nuuk]

While testing them, I found out that they do get hot.

Me again 😉 Carlos, I have had my GC monoblocks running continuously for four hours as I am trying to break in my rebuilt Goodmans 201s. When I read your post I went straight over to see if they are hot but they are only very slightly warm and it is very hot here today.

My MUR860's are just glued to a piece of fibreglass.

 

[carlosfm]

My MUR860's are just glued to a piece of fibreglass.

Without anything they get hot, but it won't burn your fingers.
What you used may be sufficient, as I said, even a small heatsink will do.
I used one of those very small and cheap heatsinks for the 78/79xx regulators for each pair of MUR860s and it works fine.

 

[GrahamnDodder]

Does anyone know if the tab where you would bolt an 860 to a heatsink is isolated?

If so, just a chunk of perforated aluminum strip would be fine. It could also serve as a structural mount to the chassis...

 

[Mad_K]

Talking about heatsinking, what material are you using for heat tranfer isolation washers ? It looks like ceramic.

if it is where did you get it from ?

cheers

mike

I you are talking about the white thick ones from an earlier pic of my BGC; they are aluminium oxide washers for to3p (to247) -These are std item that should be available by any big electronics supplier. Since I was tweaking alot I reverted to std silicone washer (There is also another reason which I won't say out loud though 🙂(

They do get hot, sometimes very hot (like when drinking cognac with friends), but they can take it 😉

 

[jan.didden]

Without anything they get hot, but it won't burn your fingers.
What you used may be sufficient, as I said, even a small heatsink will do.
I used one of those very small and cheap heatsinks for the 78/79xx regulators for each pair of MUR860s and it works fine.

If they don't burn your fingers, they're probably below 45 celcius. They can run at that temp longer than you live.

Jan Didden

 

[ALW]

I'd leave the caps off though, if it were me.

In my experience they almost always sound worse, and provide a better path for HF noise from mains supply / transformer to the PSU.

Try proper snubbers, instead, but put them at the transformer, where they belong.

http://www.hagtech.com/pdf/snubber.pdf

Peter seems to have totally misunderstood a snubber - electing to 'leave the resistor out'.

Without the resistor, it's not snubbing...

We won't touch on the star-point at the bridge instead of the caps - maybe that's history?

Andy.

 

[GrahamnDodder]

Nevermind. Figured out it is cathode.

 

[mikelm]

I'd leave the caps off though, if it were me.

In my experience they almost always sound worse, and provide a better path for HF noise from mains supply / transformer to the PSU.

Try proper snubbers, instead, but put them at the transformer, where they belong.

Peter seems to have totally misunderstood a snubber - electing to 'leave the resistor out'.

Without the resistor, it's not snubbing...

funny thing is, on simulation there is huge oscillation in the secondary after the diode swiches off, but in real life on the scope I have yet to see any visible oscillation at all.

despite this I am aware that a cap across the secondary ( I use quite a large 1uf - 4.7uf without a damping resistor ! ) brings a huge improvement in sound even though on the scope I cannot see any difference.

In theory the high value cap brings down the frequency of oscillation......even though it not visible.

could it be that the oscillation is happening in the current domain ?

 

[Nuuk]

despite this I am aware that a cap across the secondary ( I use quite a large 1uf - 4.7uf without a damping resistor ! ) brings a huge improvement in sound even though on the scope I cannot see any difference.

Another one of those arguments that is quicker to prove or disprove by trying it than talking about it!

So could I ask if that is a cap between the two secondary rails or between a single secondary and the zero volts rail. And presumably, it comes before the rectifier bridge?

 

[carlosfm]

If it has a hole, it's for a heatsink

If they don't burn your fingers, they're probably below 45 celcius. They can run at that temp longer than you live.
Jan Didden

Jan, I always think this way:
If a 78xx can't give 1A without heatsink why would a MUR860 give the 8A without any heatsink?
The same for a BUF634, don't expect them to deliver the 250ma they say on the datasheet without heatsink.
As simple as that.
I tested the MUR860s without any load and they got slightly warm, so I suppose that when you turn up the volume on your amp they may become quite hot.
It's not a matter of reliability.
If these things were made to be used with a heatsink, why not use it?

 

[UrSv]

Re: If it has a hole, it's for a heatsink

Jan, I always think this way:
If a 78xx can't give 1A without heatsink why would a MUR860 give the 8A without any heatsink?
The same for a BUF634, don't expect them to deliver the 250ma they say on the datasheet without heatsink.
As simple as that.
I tested the MUR860s without any load and they got slightly warm, so I suppose that when you turn up the volume on your amp they may become quite hot.
It's not a matter of reliability.
If these things were made to be used with a heatsink, why not use it?

I always look at it this way. When a 78XX gives 1 A with a drop of 10 V it dissipates 10 W all the time. When a MUR860 gives 8 A for a little while with a drop of 0.6 V it dissipates 4.8 W peak and significantly less constantly. When running an IGC it is even FAR less. Thus I do not feel the need for a heatsink. However, I do feel the need to mount them securely and that can be done on a piece of aluminium...

 

[carlosfm]

Re: Re: If it has a hole, it's for a heatsink

I always look at it this way. When a 78XX gives 1 A with a drop of 10 V it dissipates 10 W all the time. When a MUR860 gives 8 A for a little while with a drop of 0.6 V it dissipates 4.8 W peak and significantly less constantly. When running an IGC it is even FAR less. Thus I do not feel the need for a heatsink. However, I do feel the need to mount them securely and that can be done on a piece of aluminium...

As Mad_K said, sometimes they get very hot.
I tested them and got to the same conclusion.
UrSv, you're talking of a power amp, not a transistor radio or a preamp.
To be certain, you need to test, not just come with theories.
What makes you think your IGC doesn't need current?
I see you're one of those guys who think the power op-amps (like the LM3875) don't even need a heatsink, just put it on the bottom of a thin case and that's it.
Put Ben Harper playing at half the volume and you'll see the diodes getting hot and the power op-amps getting really hot.
And connect a pair of 4 ohm speakers to your Gainclone and you'll see how it gets hot.
NO MIRACLES HERE.

 

[ALW]

funny thing is, on simulation there is huge oscillation in the secondary after the diode swiches off, but in real life on the scope I have yet to see any visible oscillation at all.

I've certainly seen it, by triggering on the DC sawtooth and then viewing the sinusoid from the transformer on a 'scope. It can be quite high in frequency though, and may not last that long so takes some careful 'scope work to view. Unless you measure the transformer characterstics, you won't know what you're dealing with. In a good TX with little stray C, the frequency may be close to the 'scope's bandwidth?

It will vary dependant upon the real-world charcteristics of the transformer though, but as I say, Peter's implementation does little to damp any of this, but does provide a better path for the oscillation or noise (if present) into the equipment being powered. That's not what I want to do in a PSU, generally.

As for a cap sounding better, that may be - the frequency will be lower and may be more readily filtered. Surely it'as better to eliminate it entirely though - using a simple RC isn't exactly difficult.

The C part isn't critical, it just has to have low Z at the resonant frequency.

Andy.