paralleling Sanken STD03P/N (SAP16etc.)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dear All.

A question. Since the Sanken STD03 (and the former SAP15,16 etc.) has compensation diodes inside the package, the bias circuit is rather different without VBE multiplier. Makes life easier :D but what to do if you want to parallel those devices? How to manage the bias? Give every pair an own current source, and an adjustment between the diodes is not the solution I seek for. I know Arcam (P1) and Musical Fidelity use multiple pairs of those Sanken’s with only one bias adjustment pot. Stupid of me... I see those amps many times under the hood and never paid attention to it....

Who can help.. ;)

Ps: I experiment with those new STD03 devices, and I truly got wonderfull results already. I like simple designs.

With kind regards,
Bas
 
Hi Sebastiaan

normally an amplifier with multiples pairs of output transistors only have 1 vbe multiplier.

Use only the diodes of one pair of stdo3 and left the others pairs diodes disconnected (i have never use this transistors, but seems logical).

I have see the scheme of your lme49810\std03 amplifier, and i have some questions for you, if you dont mind... but later.;)

Best regards,
Sergio.
 
smms73 said:
Hi sakis

Do you mind to explain... the difference between 4 or 10 parallel devices. :confused:
the only thing that i can think off, is the thermal resistance of the heatsink.

regards


I tend to use an overly large heat sink and a pair of fans on my amps. This reduces the chances of thermal runaway.
I use a PCB case with a fan in the front and the back.
The heatsink is almost the height and width of the PC case.
 
nigelwright7557 said:



I tend to use an overly large heat sink and a pair of fans on my amps. This reduces the chances of thermal runaway.
I use a PCB case with a fan in the front and the back.
The heatsink is almost the height and width of the PC case.

Janneman from this forum did an article for Elektor (Pax Amplifier) in which he used parallel devices -- per Sanken each pair requires it's own trimpot owing to the different beta's of the pairs.
 
never read this article

but this come in to my mind and that is why i asked how many devices parallel....

crossed my mind ..... i ve seen this greek made sziklai ( with 4 sets of outs per ch.)as i said that the aythor claims that since each output has its own driver the amplifier has amazing driving abilty ....

overkill but imagine if except indepented driver was existing indepented VBE multiplier ....

i cannot even imagine if what i say has any point but just crossed my mind
 
smms73 said:
Hi Sebastiaan

normally an amplifier with multiples pairs of output transistors only have 1 vbe multiplier.

Use only the diodes of one pair of stdo3 and left the others pairs diodes disconnected (i have never use this transistors, but seems logical).

I have see the scheme of your lme49810\std03 amplifier, and i have some questions for you, if you dont mind... but later.;)

Best regards,
Sergio.


Dear Sergio,

Sorry I lost my own topic out of sight, but I am back :D Please ask me anything you want to ask.

With kind regards,
Bas
 
Re: never read this article

sakis said:
but this come in to my mind and that is why i asked how many devices parallel....

crossed my mind ..... i ve seen this greek made sziklai ( with 4 sets of outs per ch.)as i said that the aythor claims that since each output has its own driver the amplifier has amazing driving abilty ....

overkill but imagine if except indepented driver was existing indepented VBE multiplier ....

i cannot even imagine if what i say has any point but just crossed my mind


That was the intention of my experiment with multiple STD03's parallel.

Look for example at Spectral amplifiers. It are paralleled stages, with each stage has it's own driver and VBE multiplier.

With kind regards,
Bas
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.