spritzer

Yes, it's still good.

Takes all the fun out of it though...

No load for the PSU needed. It's regulated so you will see pretty much the same numbers with and without the amplifier boards present. The trimpot is a voltage divider and thus adjusts the bias supplies voltage doubler potential. All the GND connections are connected together on the ground plane on the PSU PCB. These aren't connected to chassis by default so you need to run a wire to the chassis, either direct or through a ground loop breaker. I for one always go direct. For the amp boards, each one has one wire to the PSU ground, don't connect them to the chassis as that will generate ground loops. No need to conect input or outputs when testing Also remember to power off before inserting the servo opamp. Better yet would be tom use the jumper in the position marked "servo"

For the record, I bought those Exstata board sets that were on HF. It's mostly so we can measure them and add it to the database of measured amps as I will build both of them using the same parts. The psu is garbage and that was pointed out to Alex time and time again but he's too idiotic to see reason. I sure as hell won't be using the PSU boards as the KGSSHV mini PSU is smaller and far better. Either raise the voltage rails or push more current through it. You will need new heatsinks for the transistors though.

It has more power and is more neutral so yeah, trounces it easily.

OPA445 for the higher rails.

Nothing to add a CCS too, it already has 1968's for that role. Running the output stage harder would be my go to mod but I'm not sure how much that transformer can take.

The 313 is very similar to the SRM-3 so you weren't totally off. Similar build too with the PCB at a 90° angle and stupid shit like that. A simple amp like the 323S is on my to-do list. Waiting for Kevin to have his power restored so he can check out some of my ideas.

4" would be perfect for this.

You truly don't get this place do you? We call people on their shit and claiming that changing good caps for bad is beneficial falls under that scope. You haven't added anything here that is beneficial to the community so just fuck off.

Do you even get the concept of cloning? We have improved plenty of amps and other designs over the years but this is a direct clone so it will be done 100% as Krell did it.

We want everybody stay at 350V. 400V is most certainly a "do this at your own risk" affair.

With the right tools then it is certainly possible to build this without much trouble. Stick with a good soldering iron with a small chisel tip, thin solder wire (I use 0.5mm) and taking care to only use as little solder as it possible. Also cut the leads after soldering up close to the PCB as the density of these boards is reaching silly levels so pads can be very close together but should not be connected. The main issue would be the heatsinks and how to assemble them properly. The full B+ sits on the 10M90 transistor tab so this needs to be insulated. As for the rail voltages, the amp is supposed to run at +/-350V but I usually run them at 400V. No real improvement by doing that but I've been using up my scrap BHSE transformers which can't deal with that load. Ohh and this amp eats the 717 for breakfast... Lowest I've dared to go for standoffs is 6mm and then the boards fit in a 2U chassis but I have to select what tubes to use. Parmetal might have some boxes that work better as they aren't as fixed into the 40/80/120mm sizes.

You assume that I cared enough to read your posts but I didn't.

Why on earth would we do that? This is a clone so the board will stay exactly as Krell left it.

That would be customs. Been down that road many times but USPS tracking is the worst.

Yeah, spotted the hole snafu when resizing the board over the weekend.

Less voltage swing and power aside, there are none.

Only places I've ever used the OPA2604 was in the only cmoy I ever built (good project to do after building a Blue Hawaii or two) and when upgrading the Stax SRM-300 amps. Better than the stock LF353...

It offsets my petite stature...

Post count means nothing here and it shouldn't mean anything... ever. We welcome any good engineering but swapping caps really doesn't cut it here.

This is what we use for 1.7V led: 859-LTL-307ELC Also the footprint for the resistors on this one is 0.5 and not 0.6 as on most of the other designs.

No need to get personal but Jensen caps are overpriced garbage. They are just expensive with gold printing on black which has to mean they are good. If you want to improve the SRM-323 then regulate the power supply, increase the output stage bias and add a servo. They are actually real circuit improvements which is what we do here

Well yeah and there are subtle changes elsewhere but they are small enough to be blown out of proportions by the changes in temperature.

It's a SRX based amp with probably an added phono stage. It's a good circuit but the modern ESX version is so much better.