[Uncle Bob]

A quick progress report with my build and a couple of questions

I've installed H- (LM337) and get -12.6vdc at TH which is correct according to notes on the website.

I've been a bit paranoid about heatsink to device continuity as I think that's what caused my problems initially. I think that I perhaps overtightend the heatsink bolt and caused the insulating washers to deform on the narrower part that is designed to sit just inside the mounting hole on the heatsink. Because of this, I've been a bit concerned when I measured continuity between heatsink and devic after mounting LV+. But it appears that the centre pin and the heatsink lugs are both connected to ground on the pcb - is this correct? Similarly, LV- has pin 1 connected to ground as well as the heatsink.

Just checking before I go any further.....

 

[runeight]

That's correct on the regulator pins that are grounded and the heatsinks.

 

[runeight]

Oh, reading your post again, even thought one pin on each of the regs is grounded, the metal tabs on the regs should not have continuity to the heatsinks.

 

[Arkku]

(Except LV+, as its tab is connected to the pin which is supposed to be grounded. Assuming the heatsinks themselves are grounded.)

 

[holland]

Quote:

Originally Posted by wiatrob /img/forum/go_quote.gif So while it does appear to be there - I'd never subject my ears to that level with music with those 'cans

My hiss is with the volume all the way down. disconnecting the pot and grounding makes no difference. I suspect FAs is the same, from the description.

 

[runeight]

Gents, I am still a bit stumped. Some of you are having an audible hiss, some are not, and some hear it when the vol pot is turned up with certain headphones.

I think that all of you have the new HV supply on a breadboard. And we are pretty sure that the buffer is not the source of the hiss. This leaves the B+.

And this would say that some of the PSs are noisier than others using the same component values (I assume). So why would this be so? It could be a grounding issue. That is, where do the grounds from the HV supply attach to the main board?

I don't have any further insights at this moment.

 

[Ferrari]

Me neither.

Regarding the parts I used on my perfboard HV B+ PS, well… apart from the TL783 regulator, all other parts are deviated from the BOM a bit (or a byte) since these parts are what I have in my parts bin accidentally. C8P – C11P are 220/100V on one board and 220uF/200V on the other board while C12P is 680uF/200V bypassed by 1uF MKC. The PS output filter cap C13P on the 2 boards are 10uF MKC and 33uF MKP respectively. All diodes in the PS are 1N4007.
The ground lead is directly connected to the transformer secondary center tap.

I don’t know how the PS layout of the problem case here is done? how the AC and DC traces are wired/running?
For audio applications, PS layout must be treated with utmost care and lots of dedication.
Multiple crossing and paralleling AC and DC traces (for example) can induce noise in the circuit, which undoubtedly will be audible at the amp output.

 

[holland]

First off, my sincerest apologies for not being able to investigate this further. I seem to have misplaced my 400MHz probes and my 100MHz probes are broken.

I'll keep searching for them. I'll see if I can find someone with a scope...I'll bring it to work if I have to, but that's sort of frowned upon. AMB is near me, perhaps I can give him a buzz, if he's not too busy.

I checked with the Philips SBC-HP1000 headphones, no hiss. It's very suspicious. The last time I saw something like this (though it wasn't a hiss), the buffers were oscillating due to the inductive load of very specific headphone cables. A zobel on the output jack was the solution.

Here are my pictures. I moved the black wire. It used to be on top, running parallel to a diode. I figured I'd move it underneath with some space and see what happens. No change.

The black wire is ground and goes to the xformer through the 3-terminal jack.

also, yeah, busted, I didn't clean the flux off of it.

The 3rd picture, I changed the routing of the wires. I didn't run wires under the PCB, but underneath the chunk of aluminum to the left. I also had some LEDs wired under the tubes from the proto phase to determine if it was on or not...those have been removed.

 

[runeight]

No apologies necessary at all. I am just grateful for the time you have spent already on this project.

Here's one experiment. Try running the ground of the HV board directly to the star point instead of somewhere else.

 

[rhester]

If I were to build this as a 2 box project (transformer in a separate box), how would grounding go? Still star ground the pcb to frame with screw / standoff, and in 2nd box tie incoming ground to chassis? Any benefit besides helpin gtake noise out of the question?

And on signal input, how has everybody been running to that front input location? IS there a need for shielded, and if so, where do you ground the shield ?

 

[runeight]

Quote:

Originally Posted by rhester /img/forum/go_quote.gif If I were to build this as a 2 box project (transformer in a separate box), how would grounding go? Still star ground the pcb to frame with screw / standoff, and in 2nd box tie incoming ground to chassis? Any benefit besides helpin gtake noise out of the question? And on signal input, how has everybody been running to that front input location? IS there a need for shielded, and if so, where do you ground the shield ?

Yes on the grounds. You may also want to include a means of connecting the chasses of the two boxes. You may or may not actually need this. However, if you run it in the cable where the trafo wires go, there is a risk of picking up hum in the grounds.

MHO is that it's always best to shield the input wires unless the run is very short. In this case, the shield should be grounded at the end where the wires connect to the input terminal block on the board. There is a ground terminal there for this. The jacks should be insulated and the shield should also be connected to the grounds of the jacks at the other end of the wire.

 

[holland]

going to star ground with a solid 30awg wire seems a bit quieter, the shape changes as well. From a SSSSS to a zzzzz. I would guess (not measured as no eq to do it) about a 3-5dB drop (ears not so good a measurement device). The hiss to buzz could be the inherent noise floor of my environment. I know if I max the volume on any of my cans, I hear a buzz.

Original was an 18awg solid core to transformer center tap. EI core wallwart with no earthing (3rd prong not connected). I could try to put that to the center tap, it's a pain since the wallwart enclosure is so small.

 

[wiatrob]

I'm using a homemade PCb per KKlee - Grounds from Phoenix connectors straight into the Transformer in connector. I've only removed about half the old PS components (C10 and 3 diodes), HV PS feeds in thru J3.

I just noticed that my star ground appears to have loosened... I'll fix that when I return from travel next week...

 

[bmwpowere36m3]

I have a few questions regarding the SOHA II build:

1. I'm assuming Ferrari's revised HV PS has been implemented in the production board, correct?

2. Are there any modifications nessecary to the "stock" HV PS to run a B+ voltage of ~100V or around there?

3. Using the most current BOM of Cavalli's site (7007b), are there any useful modifications to be made?

4. Finally what components would benefit most from 'boutique' stuff, i.e. C1L/R to use a nice PIO film cap?

 

[runeight]

The SOHA II that is implemented on the board is shown on the website under SCHEMATICS. This implementation, with the trimpot, should give you a range of B+ from about 55V to 95V. Maybe a bit higher than 95V.