Paradise Builders

sq225917 · 2013-07-22 12:51 pm

amsoque said:
Also, 26Vdc is the MINIMUM recommended input voltage for the PSU.

Hi Frans,

in a previous post it was stated that the input voltage have to be between 25 and 26 V. Now, you say minimum 26 and SQ have 28 V. That add quite some heat to the power transistors.

So, it's better with 26-27 volts?[/QUOTE]

I have 28v off the transformers, but only 26.5v coming out of the power supply to the shunts. My boxes are built as a folded 'U' shape sled which slides into an extruded sleeve. They aren't designed with any ventilation, but there is a 1mm gap between sled and sleeve that runs the entire length of the upper edge of the front and back panels.

FdW · 2013-07-22 1:52 pm

sq225917 said:
[/COLOR]

Hi Frans,

in a previous post it was stated that the input voltage have to be between 25 and 26 V. Now, you say minimum 26 and SQ have 28 V. That add quite some heat to the power transistors.

So, it's better with 26-27 volts?

I have 28v off the transformers, but only 26.5v coming out of the power supply to the shunts. My boxes are built as a folded 'U' shape sled which slides into an extruded sleeve. They aren't designed with any ventilation, but there is a 1mm gap between sled and sleeve that runs the entire length of the upper edge of the front and back panels.

O.K. I am wrong 25Vdc is the minimum, 26Vdc gives a bit of headroom. See Post#155 and Mpp#6462 (sorry to have created some confusion, my only excuse is I do not have my notes here at this location

). P.S. for the 25Vdc you need to check the PSU input voltage with an scope (to check for the minimum!).

These are the most important/relevant messages regarding PSU and Pre-Reg.

Have a look at these and their following messages.

sq225917 · 2013-07-22 3:03 pm

Apologies for the missed 'quote' in my previous post...

audiomania · 2013-07-23 6:08 pm

Abraxas: yeah, thought so that the heat could be a problem. But venting the case would bring you the drifting of the inputtransistors, so that they actually should be thermocoupled... How do you see that?

Abraxas336 · 2013-07-24 7:33 am

My line preamplifier is not DC coupled so I'm not too nervous about a little bit of offset drift. If you plan to stack boards and think it might be an issue you can use a "Streichholzschachtel" or another small box to protect the amplifier transistors. Or even only like a little "wall" between the Regulator and the amp.

Regards
Sven

hesener · 2013-07-24 8:09 am

A little copper foil strip glued to the input stage will do the trick alright. Absolute temperature is not a problem (as long as its not red-hot

) but temperature difference between the transistors is what seems to cause that drift.

amsoque · 2013-07-24 6:44 pm

There will be a group buy for the Calvin Buffer including the parts or only for the PCB? I'm going to buy something from Mouser and I think to add the Calvin parts to the chart.

Also, is the schematic and the BOM definitive or do you expect to change something?

Regarding the transistors drift, yes, in my case in each mirror group there is a transistor that is warmer than the others. Perhaps with the new group of Diotec matched transistors that I'm going to sort (which are by the way the best choice regarding the hfe match), this effect will diminish.

Thanks.

Calvin · 2013-07-24 6:54 pm

Hi,

...I think to add the Calvin parts to the chart

I´m not dead yet. I´m still in desperate need of my ´parts´.
Leggo You butcher.

jauu
Calvin

Benedetto · 2013-07-24 7:21 pm

sq225917 said:
[/COLOR]

Hi Frans,

in a previous post it was stated that the input voltage have to be between 25 and 26 V. Now, you say minimum 26 and SQ have 28 V. That add quite some heat to the power transistors.

So, it's better with 26-27 volts?

I have 28v off the transformers, but only 26.5v coming out of the power supply to the shunts. My boxes are built as a folded 'U' shape sled which slides into an extruded sleeve. They aren't designed with any ventilation, but there is a 1mm gap between sled and sleeve that runs the entire length of the upper edge of the front and back panels.[/QUOTE]

Hello Sq,
got some listening impressions from our man in Aire.
He told me,
after trying some tricks, (Akkus, some more Volts, some less... you name it, he did it) a Volt more or less is not vital.
I think Alfred will figure out one or another trick one day.

So far you can cool down if your PSU spits some 26Volts.

amsoque · 2013-07-24 7:43 pm

Hi Calvin,

I feel I'm bothering you, but I can't understand exactly why and how. Perhaps is my bad english that doesn't help me to understand the subtlety of your messages. Would you like to tell me clearly what is the problem? And I take the permission to ask you to address me directly, frankly, without paraphrases and quotes. Unfortunately, I don't have a good use of English.

Emil

hesener · 2013-07-24 7:55 pm

Benedetto said:
I think Alfred will figure out one or another trick one day.

Yes, that man experiments a lot and keeps me updated on what he does - very good stuff. He promised me to post his experiences soon....

I will now reveal one possibility we discovered together: Shortening one of the LEDs in each leg of the shunt will reduce the supply voltage to 16.5V. But the input voltage for the CCS can be reduced to 24V - voila, supplying with batteries is now possible, which has a couple of sound implications that sound really nice (havent tried it myself yet, but some day I will)

stay tuned......

hesener · 2013-07-24 7:57 pm

amsoque said:
There will be a group buy for the Calvin Buffer including the parts or only for the PCB? I'm going to buy something from Mouser and I think to add the Calvin parts to the chart.

Also, is the schematic and the BOM definitive or do you expect to change something?

Hi Emil (and all), please wait one more day, I hope to post the document on the Calvin Buffer for Paradise soon. It works fine and uses mostly standard non-SMD parts, just a question of documentation. This will contain a BOM so all you guys can start ordering components.

And yes, there will be a PCB group buy very soon....

RCruz · 2013-07-24 8:41 pm

Calvin said:
Hi,

I´m not dead yet. I´m still in desperate need of my ´parts´.
Leggo You butcher.

jauu
Calvin

amsoque said:
Hi Calvin,

I feel I'm bothering you, but I can't understand exactly why and how. Perhaps is my bad english that doesn't help me to understand the subtlety of your messages. Would you like to tell me clearly what is the problem? And I take the permission to ask you to address me directly, frankly, without paraphrases and quotes. Unfortunately, I don't have a good use of English.

Emil

Hi Emil

Calvin is just having fun .... He means he is still alive so he wants to keep "his parts"

You are not bothering anyone... be assured.

amsoque · 2013-07-24 9:01 pm

Thanks Alfred. Appreciate.

hesener · 2013-07-25 2:58 pm

Calvin Buffer for paradise

Folks, here is the assembly guide for the Calvin buffer, and I will start the group buy over in the group buy section very soon - waiting for a quote. Hope you like it!

hesener · 2013-07-25 3:14 pm

Group buy for the Calvin buffer for Paradise is on:

http://www.diyaudio.com/forums/group-buys/239776-calvin-buffer-paradise.html#post3573058

miepa · 2013-07-28 11:45 am

@alfred

1) should the transistors in the calvin buffer be matched?

2) the foilcaps could also be SILVERMICA or STYROFLEX or better go with Foil?

thank you!

Calvin · 2013-07-29 7:08 am

Hi,

The master JFETs Q1 and Q4 should be matched, as they'd be responsible for the most of the current imbalance of upper and lower part.
The cascoding JFETs Q2/Q5 may be screened for gross Vgs variations, but don't need matching.
Their Vgs is at the same the master's Vds.
A Vgs of min. V up to 5V should establish (this holds true for the 4391 JFETs).
The bipolars may be used off of the shelf.
I'd preferred a layout with closer proximity of the master JFETs to allow for thermal coupling, even though Joachims test showed low sensitivity to thermal instability.

C3 can be a good quality film cap, as may be C1/C2.
You won't get a silver Mica with 220nF.
If you can accomodate for larger values You may increase C3 up to 1uF.
C1 and C2 may be silver Mica, Styro, or ceramic COG/NPOs, soldered closely to the transistors legs.

jauu
Calvin

miepa · 2013-07-29 11:03 am

thank you Calvin!

Werner Hammelmann · 2013-07-29 3:52 pm

A little copper foil strip glued to the input stage will do the trick alright. Absolute temperature is not a problem (as long as its not red-hot

j2xgwDMBDSsnVA69OtCogRQzID4UOTOFEwAHkGJSmhPzgRIAFU7lagXmBJGuXwEwEMPlz8AaBEagAjDrBQAkcUj5iCtQwK0FSzhciBsQADs=

) but temperature difference between the transistors is what seems to cause that drift.

Alfred is absolutly right

A small copper plate 1mm works very well.
I pasted it with a very little drop of silicon on top of the transistors (see fotos)

Outlet drift before
Right cannel -3,5 / +10,5 mmVolt
Left cannel -1,5 / + 8,5 mmVolt
With copper plate
Right Cannel -1,0 / + 2,5 mmVolt
Left cannel 0,5 / + 2,0 mmVolt
That up and down is much more stabil as before, it's the most time zero
Soundwise is it a littel but only a little bit more stabil and the stage a bit wider. Thats my impression.
The sound before was very good and nice as well, but you can hear a small differens. My MK 4 / R. Hoffmann with (MC 10 Menno Van der Veen) step up transformers , had it not so easy any more. At the moment I like and hear both of them, in addiction to the records

In a copple of days I will give some impressions about my batterie trails,
very interesting as well.
Next step up is then the "Calvin Buffer" Thanks Calvin and Alfred.

Take care
Werner

javascript:;

Paradise Builders

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Attachments

Member

Member

Member

Member

Member