Alternatives to Sanken SAP15 in an Arcam A65

[JonHarrison]

Hi,

I've got an Arcam A65 with blown output transistors (Sanken SAP15N/P).

I was thinking of modifying the amplifier anyway and now I'm wondering what to replace the output transistors with :-

Another set of Sanken SAP15N/P
Sanken STD03N/P
Onsemi ThermalTrak NJL3281D/NJL1302D with additional drivers
Something else ...

Has anyone done anything similar already ? What are your thoughts / suggestions / recommendations ? Let the debate begin !

Cheers,

Jon

[Sebastiaan]

Dear Jon,

If I was you I would go for the STD03. The design and current sources are designed for the SAP and STD devices. Don't forget to add a emitter resistor, since the STD03's don't have them inside anymore.

With kind regards,
Bas

[JonHarrison]

Hi Bas,

Thanks for the reply. 0R22 or is it worth experimenting do you think ?

Cheers,

Jon

[jaycee]

The weak point of the SAPs was that the onboard emitter resistor would fail long before the transistor die. The STD devices are simply the SAPs without the resistor! I'd stick to the same 0R22's personally.

The weak point of these amps IMO is the stupidly small heatsinks. In the service manual, it's worth noting that they reduced the bias on the A65+/A75+ - it was 37mV on the non-Plus versions, and is 12-15mV on the Plus!

One of the good points on this amp though, is the use of a voltage doubler to create a regulated power supply for the frontend circuitry in the power amp, seperate to the supply used for the output stage. I'm not sure about the Darlington arrangement in the LTP - personally I think they should've gone for a CFP arrangement if they were going to go to the effort!

[Sebastiaan]

Originally Posted by JonHarrison

Hi Bas,

Thanks for the reply. 0R22 or is it worth experimenting do you think ?

Cheers,

Jon

Dear Jon,

The STD03's feel happier with 0.22 Ohm. I experiment with lower values (0.1 Ohm) but with a slight bias miss adjustment you get thermal runaway. So like jaycee says, stick with 0.22Ohm which seems a good trade off between reliability and sound quality.

With kind regards,
Bas

[Sebastiaan]

Originally Posted by jaycee

- it was 37mV on the non-Plus versions, and is 12-15mV on the Plus!

37mV is far above what the Sanken application note suggested. As an matter of fact with idle bias above 100mA they can't guarantee stabile thermal behavior anymore, since the diodes are designed for 40mA.

Does one one have service manuals from the A65+/A75+ and non plus?

As return I have service manuals from the P1 power amp (which use2 pairs of SAP output devices), the A60 the the Alpha 10 and the Alpha 9.

Thanks in advance.

With kind regards,
Bas

[JonHarrison]

Hi guys,

Thanks for the info.

I'll see if I can mount some STD03 on a larger heatsink. The heatsinks in the Alpha 10 were tiny too (Quasi Mosfet) - I haven't seen the inside of any later Arcams. I haven't done the calcs but I'm guessing they're not sized for prolonged high output levels.

Yes not sure about why they chose to increase the gain in the first LTP and then have a second one.

I'm not sure there was a seperate service manual for the + versions. I have the A65/A75 manual and it mentions the bias settings for the A65/75/85/65+/75+. Bas, send me a PM with your email address and I'll forward it to you.

Mods - I was thinking Alexander current feedback amp like some of the other Arcams with the Sanken transistors.

Cheers,

Jon

[Sebastiaan]

Thank you Jon, you have a PM. If I have your email address, I will sent you the official Sanken SAP application note, which is hard to find. This application note is very useful for understanding the devices and the recommended setup and bias settings.

With kind regards,
Bas