Elektor's "Ulti Amp"

[OliverD]

Has anybody tried what at least in the German issue was called the "Ulti Amp"? It was published in 5/95. Measured specs were more than excellent, but I'm interested if anyone can comment on the sound.

The dual transistors used in the two gain stages (2N2914, BFX36) have long been obsolete anyway. I've always wanted to build one (some parts are still in my junkbox), but never proceded as I came across other concepts (current feedback, class A) I found more interesting.

The protection circuit is somewhat special (it shorts the PSU rails via Triacs instead of opening a speaker relay) and can easily be adopted for use in any amplifier.

[blmn]

AMT-freak,

Can you share the schematics with us?

Tks in advance.

[OliverD]

Sure.

Note that this design was an "update" from a high-current audio amplifier for very low impedance speakers, using the same PCB. Irritating traces of those modifications are left both in the schematic and PCB layout.

Here is the main circuit without power supply and protection circuit. The optocouplers are used as an electronic switch instead of a speaker relay.

[OliverD]

Protection circuit. Optocouplers are connected at B1 / B2. The Triacs short the supply rails to trigger the secondary fuses.

The circuit was published in 5/94, not 5/95 as I wrote in post #1.

[Piersma]

HI AMT-FREAK,

I have build four of these mono blocks. The sound is detailed, clear and the bass is extremely controlled. There are some serious problems with the building of the amplifier. I can give you details if you want to. I redesigned this amplifier, it is almost ready.

I have some obsolete items for instance the matched transistors.

Regards

[OliverD]

Hi Piersma, thank you very much.

I won't build this circuit in the near future (too many other projects on my list), but a friend of mine started this project and since he has no experience in electronics, I promised to help him. We did the original PCBs at home, but building the circuit on them was a mess. You probably found the same errors as we did .

Anyway, I've started a new PCB layout (single sided this time), but we were both quite busy in the last few months and didn't go on with the project.

My friend might be interested in buying the original dual transistors from you, please email me for details. (We hand-matched similar single transistors and thermally coupled them.)

What other amps did you compare the sound to?

AMT

[halojoy]

This one uses "my favourites" MJE15030 MJE15031
(best commonly availeable TO220 there is)
as both
Voltage amplifiers VAS and
PreDrivers for the output Devices.

Output uses 2SC2922 BIPOLAR.

/halo - Elektor subscriber - has the article

[edm]

There are some serious problems with the building of the amplifier. I can give you details if you want to. I redesigned this amplifier, it is almost ready.

Please share those issues, I've ordered today most of the parts for an amp which is based on the ultiamp. The etching of the modified schematic is due somewhere this week, so if i have overlooked something it is not too late to repair...

Thanks,

Emiel

[Piersma]

Hi Emiel,

First of all the BFX and 2N matched transistors can be replaced by the MAT02/03 or the SSM2210/2220. The matching of the MAT and SSM types is better than the BFX/2N. They can handle the Vce is this schematic.
Because of the high gain in the first stage, use 0.1% collector and emitter resistors. Otherwise use offset trimming.
BFX and 2N collector currents should closely match !
The FET current sources (BF256) give an unprictable current because of variations in IDSS. Usource should be higher than Ugate.
Current sources BC141/161 should extended with heatsinks.
They get extremely hot because of the high Vce.
The MJE15030/31 (T12/13) should also be extended with proper heatsinks they also get extemely hot.
The Zobel network R47/C12, don't use a wirewound resistor but a metalfilm. (3W type)
The function of R51/C13 is unknown to me because of the proposed connection.

Regards

[diyman]

Would the operating voltage of the SSMs a bit too low for comfort? I believe the absolute max Vceo of the SSMs are about 40V.