Inside High End Equipment

[elcorso]

3 hours ago, Sam Lord said:

A stereo microphone preamplifier with a single power transformer, yet with *unmeasurable* crosstalk: < -140dB at 200kHz. Try to match that.

 

Beautiful and sharp Sam.  But now we need George approval for the case 

 

Roch

[marce]

9 hours ago, semente said:

 

What about ventilation, isn't it an important requirement to ensure stable performance and longevity of components?

Using the case as a heatsink is common way of extracting heat from the electronics... 

[Sam Lord]

5 hours ago, elcorso said:

 

Beautiful and sharp Sam.  But now we need George approval for the case 

 

Roch

 

I'm sure George would approve, it's just bolt-together aluminum, though very precisely machined.  That world-leading unit costs just $3200 with remote gain control (signaled using a separate XLR cable).  But I should have added the designer and hand-builder of every single unit: Grant Carpenter of Gordon Audio.  He has actually measured the effects of steel vs. aluminum chassis for his circuits because they are so sensitive.  In fact, in completed units many years ago he detected hysteresis artifacts of stainless steel vs. magnetic machine screws that bind the chassis, so he went for stainless steel screws throughout.  Characteristically, he shared his findings with some of his competitors to help their performance.  The high-end community in professional audio is a very tight and dedicated bunch.

[Sam Lord]

12 hours ago, semente said:

 

What about ventilation, isn't it an important requirement to ensure stable performance and longevity of components?

 

2 hours ago, marce said:

Using the case as a heatsink is common way of extracting heat from the electronics... 

 

Ventilation certainly helps longevity, even with massive heatsinking.  But for components using less than about 30 watts in a full-sized (1RU or bigger) unit, I wouldn't worry.  A bigger concern is with internal parts which lack bolt-on heatsinks.  I've seen many parts like TO-220-cased regulators and transistors that have no business operating without heatsinks.  A basic LM317 regulator is rated about 15 times higher dissipation when well-sinked compared with unsinked operation. 

[gmgraves]

18 hours ago, semente said:

 

What about ventilation, isn't it an important requirement to ensure stable performance and longevity of components?

Of course ventilation is important, but it doesn't require several thousands of dollars (at retail level) worth of CNC machined aluminum billet to get adequate ventilation. That can be just as easily accomplished with holes punched in cheap sheet steel!

[Ralf11]

maybe they used the aluminum billet as a neutrino shield

[One and a half]

12 hours ago, Sam Lord said:

 

 

Ventilation certainly helps longevity, even with massive heatsinking.  But for components using less than about 30 watts in a full-sized (1RU or bigger) unit, I wouldn't worry.  A bigger concern is with internal parts which lack bolt-on heatsinks.  I've seen many parts like TO-220-cased regulators and transistors that have no business operating without heatsinks.  A basic LM317 regulator is rated about 15 times higher dissipation when well-sinked compared with unsinked operation. 

A regulator fitted to a grounded heatsink is an open invitation for high impedance leakage currents to form. 

[sandyk]

17 minutes ago, One and a half said:

A regulator fitted to a grounded heatsink is an open invitation for high impedance leakage currents to form. 

 

Adhesive Silicon Washers or Mica washers should reduce this .

I became aware of this problem when I made a + 12V 2A and+5V 2A PSU for HDDs etc. using MC78T05 3A and MC78T12 3A regulators where they were bolted to a large common heatsink on the rear aluminium panel of the case.

 Other than  the 78XX , most adjustable voltage regulators need to use washers as the metal tab is not the 0 volts connection. 

[One and a half]

11 minutes ago, sandyk said:

 

Adhesive Silicon Washers or Mica washers should reduce this ? .

 Other than  the 78XX , most adjustable voltage regulators need to use them as the metal tab is not the 0 volts connection. 

My experience is with 300A+ Power transistors. Even though the base is metal, the interior silicon stack forms a capacitor between the substrate and the metal. This propagates RF when the transistor switches in a PWM fashion about 1kHz and under. For HV applications 3300 V-11,000 V, the heatsink floats above ground to avoid common mode noise output. Mounting is via ready made heatsink wafer or a thin layer of silicon paste applied with a paint roller.

 

This is an older 3rd gen IGBT.

 

 

2MBI400VD-120-50.PDF

[One and a half]

On 1/27/2018 at 7:37 AM, sandyk said:

 

How about something like this?

Incidentally, the John Curl Blowtorch Preamp part 2 Thread in DIY Audio , now has almost 100,000 replies !!!

 

Click on the image twice for a much larger image.

 

That's a Lumin xx1 series. There's minimal power losses in that design, 25W , tops, so the heat is dissipated through the solid aluminium easy enough.

[sandyk]

3 minutes ago, One and a half said:

Mounting is via ready made heatsink wafer or a thin layer of silicon paste applied with a paint roller.

 

 My mind boggles at the thought of applying heatsink paste with  a paint roller  !

 

Are Aussies the only ones who don't put a space between " heat"  and "sink" ?

 I often  do to avoid nit picking by the U.S. members.

[sandyk]

2 minutes ago, One and a half said:

That's a Lumin xx1 series. There's minimal power losses in that design, 25W , tops, so the heat is dissipated through the solid aluminium easy enough.

 I corrected that later.

 I later found a genuine photo of the Blowtorch, but didn't post it, as I can't see why so many DIY Audio members still keep posting about such an uninspiring looking item, with only 40 ever made.

[One and a half]

1 minute ago, sandyk said:

 

 My mind boggles at the thought of applying heatsink paste with  a paint roller  !

 

Are Aussies the only ones who don't put a space between " heat"  and "sink" ?

 I often  do to avoid nit picking by the U.S. members.

I use heatsink, spell checker hates it, never mind!

 

For routine maintenance, the Xeon workstation's heatsink paste is removed with WD40 and a rag, let to dry. Bunnings have a very small paint roller, it's only about 20mm wide, just perfect, takes a bit of effort to get the paste onto the roller, but you can't beat it for providing an even amount of paste. Just paint on until the top of the processor until the grey is gone, done!

[One and a half]

1 minute ago, sandyk said:

 I corrected that later.

 I later found a genuine photo of the Blowtorch, but didn't post it, as I can't see why so many DIY Audio members still keep posting about such an uninspiring looking item, with only 40 ever made.

Nice! Dual mono and the PSU are out of the box, grand idea!

[elcorso]

12 hours ago, gmgraves said:

Of course ventilation is important, but it doesn't require several thousands of dollars (at retail level) worth of CNC machined aluminum billet to get adequate ventilation. That can be just as easily accomplished with holes punched in cheap sheet steel!

 

I hate ventilation holes because living in the Rainforest there is too much dust in the environment. It does not matter how sealed your doors and windows are. Not to mention insects and other things ...

 

The only picture I published on this thread is the Conrad Johnson ET-5 preamp, SE Class A SET. No heatsinks and no vent holes on his steel chasis. Only one tube and a lot of capacitors, so it does not need extra ventilation, but it does reach the optimal operating temperature without overheating.

 

Roch

[semente]

16 hours ago, gmgraves said:

Of course ventilation is important, but it doesn't require several thousands of dollars (at retail level) worth of CNC machined aluminum billet to get adequate ventilation. That can be just as easily accomplished with holes punched in cheap sheet steel!

I agree.

[gmgraves]

1 hour ago, semente said:

I agree.

Although this new John Curl designed Parasound Phono Preamp is not housed in a case machined from a solid billet of CNC machined aluminum (but the front panel is as is the partition around the main circuit board), It sure doesn't give you much for the $2300 asking price! Compare this for $2.3K with this for (roughly) the same price. Somebody is ripping-off somebody...  

[Superdad]

1 hour ago, gmgraves said:

Although this new John Curl designed Parasound Phono Preamp is not housed in a case machined from a solid billet of CNC machined aluminum (but the front panel is as is the partition around the main circuit board), It sure doesn't give you much for the $2300 asking price! Compare this for $2.3K with this for (roughly) the same price. Somebody is ripping-off somebody...  

 

You are not taking into account the fact that Parasound's retail prices are based on a traditional deal/distributor model.  So USA dealers purchase at about 40% off, and international importers pay get anywhere between a 50-60% discount from retail.

 

So Parasound sells that $2,300 retail product for between $920 and $1,380 to their distributors and dealers.  Thus their profit has to come out of that much lower figure.

 

Schiit Audio sells direct, so there are no dealer/distributor margins to account for. 

 

 

[gmgraves]

3 hours ago, Superdad said:

 

You are not taking into account the fact that Parasound's retail prices are based on a traditional deal/distributor model.  So USA dealers purchase at about 40% off, and international importers pay get anywhere between a 50-60% discount from retail.

 

So Parasound sells that $2,300 retail product for between $920 and $1,380 to their distributors and dealers.  Thus their profit has to come out of that much lower figure.

 

Schiit Audio sells direct, so there are no dealer/distributor margins to account for. 

 

 

You are correct, Mea Culpa. But on the other hand there is more than 2-1 difference between the dealer price of the Parasound and the price of the Schiit Yggy. Wouldn't you agree?

[PhilR]

Why should the electronics have all of the fun in this thread?

 

Magico Q5:

 

 

[Sam Lord]

On 1/28/2018 at 4:10 AM, One and a half said:

A regulator fitted to a grounded heatsink is an open invitation for high impedance leakage currents to form. 

 

I don't think it would be strong with an LM317.  We didn't ground those sinks anyway, though most probably do.  But thanks for your story on those megatransistors.  

 

On 1/28/2018 at 4:26 AM, sandyk said:

 

Adhesive Silicon Washers or Mica washers should reduce this .

I became aware of this problem when I made a + 12V 2A and+5V 2A PSU for HDDs etc. using MC78T05 3A and MC78T12 3A regulators where they were bolted to a large common heatsink on the rear aluminium panel of the case.

 Other than  the 78XX , most adjustable voltage regulators need to use washers as the metal tab is not the 0 volts connection. 

 

Yes, I haven't used devices that could be directly cooled except by a floating sink; we always insulated the tab anyway, you don't want gobs of metal in your circuit.

 

On 1/28/2018 at 4:37 AM, One and a half said:

My experience is with 300A+ Power transistors. Even though the base is metal, the interior silicon stack forms a capacitor between the substrate and the metal. This propagates RF when the transistor switches in a PWM fashion about 1kHz and under. For HV applications 3300 V-11,000 V, the heatsink floats above ground to avoid common mode noise output. Mounting is via ready made heatsink wafer or a thin layer of silicon paste applied with a paint roller.

 

This is an older 3rd gen IGBT.

 

 

2MBI400VD-120-50.PDF

 

On 1/28/2018 at 4:43 PM, gmgraves said:

Although this new John Curl designed Parasound Phono Preamp is not housed in a case machined from a solid billet of CNC machined aluminum (but the front panel is as is the partition around the main circuit board), It sure doesn't give you much for the $2300 asking price! Compare this for $2.3K with this for (roughly) the same price. Somebody is ripping-off somebody...  

 

Sure, but you aren't accounting for parts quality.  If the Parasound had bulk-foil resistors and polystyrene caps, it would have been a bargain.  But as you said, it really isn't: I don't expect the Parasound to have better devices than the Schiit.

 

On 1/28/2018 at 11:45 PM, PhilR said:

Why should the electronics have all of the fun in this thread?

 

Magico Q5:

 

 

 

Geez, why not just pour a big slurry onto the inner walls and call it good?  So much bling machining here; I'd rather have the mass and inelasticity of concrete with a good filler.  

[STC]

Dutch made hi end Sphinx Project Twelve of Siltech mono amp. Look at the tyre balance weight on the transformer for damping. 

 

 

[marce]

I would have used something a bit more up market for damping if it was required....

[Fokus]

Sphinx Project Twelve ... that is at least from 20 years ago, probably even before Siltech was founded.

 

[Speedskater]

On 1/28/2018 at 4:10 AM, One and a half said:

A regulator fitted to a grounded heatsink is an open invitation for high impedance leakage currents to form. 

There is nothing high impedance about  a LM317 regulator circuit.