The Wire LPUHP 16W Power Amp parts GB

[hochopeper]

I wonder if specifying 230V primary rating is adding to that effect. Those on the North American continent may not have the same problem, not sure how tight their distribution voltages are regulated, I expect its probably quite well regulated since they have a lot more small distribution transformers and short LV network by comparison to how grid is operated here.

Generally speaking the distribution network here is set up for 240V +/- 6% but ~260V is certainly not unheard of ... after reading this I measured line voltage at the wall on the weekend and got ~249V late Saturday afternoon.

Like you and opc have said the power supply filter caps shouldn't have too many issues with being run above rated voltage for a short period, certainly worth being aware of though!

[nattonrice]

Originally Posted by opc

do you guys actually call them crocodile clips down there?

Only the real odd balls amongst us

[qusp]

Originally Posted by hochopeper

I wonder if specifying 230V primary rating is adding to that effect. Those on the North American continent may not have the same problem, not sure how tight their distribution voltages are regulated, I expect its probably quite well regulated since they have a lot more small distribution transformers and short LV network by comparison to how grid is operated here.

Generally speaking the distribution network here is set up for 240V +/- 6% but ~260V is certainly not unheard of ... after reading this I measured line voltage at the wall on the weekend and got ~249V late Saturday afternoon.

Like you and opc have said the power supply filter caps shouldn't have too many issues with being run above rated voltage for a short period, certainly worth being aware of though!

i'm aware we are set up for 240 for the most part and we discussed that, I actually dont see it as a real problem, but as you say best to be aware of it so I posted it here. my voltage at the wall couldnt have been too high though, wasnt up for doing that at 5am or whenever it was. with a no load specified voltage of 20.7vac we get 29.187vdc before the diode drop, which with no load isnt likely to have been more than 2v I wanted the slightly higher voltage when I specified 16v secondaries instead of 15, despite what you have told me about new solar going online etc, I up till this point have found the voltage to be LOW under times of high load and that I mostly have not in the past gotten the secondary output I was hoping for when I spec 240, risking only getting 15 out of this 16v transformer....

if I were to change anything it would be the caps

Originally Posted by nattonrice

Only the real odd balls amongst us

moi? hey look i'm not the one with the urushi painted line transformers hehe

[hochopeper]

Don't get me wrong, I'm not concerned and I agree the tx spec is pretty much ideal! Agree that if anything the caps would be the place to tweak but not worth it for something that's only a concern while testing the regulators.

My point was more that on a more tightly regulated distribution network their may be less voltage fluctuations and they might not exceed the capacitor voltage spec by as much. For those in northern hemisphere that are part of the GB that is.

[qusp]

no worries yeah maybe not, but it will still go just over, I would think even the BOM transformer will.

it definitely should not should not encourage people to not build and test the regulators first, the only way for most of us to replace those buffers is by snipping the legs and thats not a cheap mistake; I suppose if people have any of that chip flick low melt solder that'll do the trick too, but I reckon that stuff is best avoided. just load the output of the reg section with a power resistor, problem solved.

Also just a few build tips, I suggest installing the 10uf ceramics decoupling the buffers supply on the 2 buffers closest to the regulators, before the PSU caps and heatsinks go in; not the buffers just the caps as they are somewhat obstructed after these go in. you can do all of them if you wish, but against my normal method of lowest to highest profile parts, this layout means I found it easier to install the buffers first thing after the PSU testing. the speaker outs can go in last unless you know exactly the length you'll need, or the wire you are using is not expensive. at first I thought this would be a problem but its really not, any speaker wire will be stiff enough that you can just guide it down into the hole through the gap in the reservoir caps.

[wirewiggler]

Not that I am in any big hurry all I have purchased is the passives and transformers but any word from hypertune on the heatsinks?

Bill

[hochopeper]

Shortest answer is, unless qusp has heard something today then, no we haven't heard from him since he last posted in the thread.

Obviously something has come up that has him unable to get online, hopefully all is well with him!

In terms of way forward from here for rest of GB logistics the following is my understanding and I hope qusp corrects me if I've got anything wrong here. To minimise double handling of kits I think the plan is to pack all of the kits after the keratherm arrives so for those without heatsinks and just wanting passives that's really the holdup at the moment and the keratherm shouldn't be more than a week away from what I know (its in transit at the moment, not sure if qusp has tracking on it) after that its down to qusp and I to find a time to get together and cut keratherm to size/pack the parts for everyone.

For the heatsinks at some point we'll all need to make a decision on that too if we don't hear from hypertune in the next week or so.

[bcg27]

Worst cast the original sinks from Digikey in the BOM should work ok right?

[qusp]

yeah, i'm actually looking at other ideas to go along with them, from my experiments we would be well served to look at sinking the underside, i'm looking at copper stock, again actually here in oz is probably going to be one of the best places to source that. I think bolting a piece (perhaps a thick angle) of copper or alloy top and bottom, with a bolt through both pieces and the pcb, with a narrower piece on the top side than the bottom and leave it at that, the tops f the buffers are really not the best place to try and take heat from, they hardly get hot at all. the one in the BOM IMO needs to be accompanied with a strip down each side. i'll model something up tomorrow if I can find the time. these are a nice idea that could work (flat cool pipes)

anyway check the build pics linked above

[qusp]

its possible that as I didnt have the input shorted there was some oscillation, but I dont think so as I had nothing on the output, the output showed max 4mv DC even with the inputs floating. it was by no means a really worrying heat, but still something I would like to take care of as it was done in the middle of the night and its nearly winter here and it was getting a little toasty under there, perhaps 50-55c or a touch more on the underside of the board and the lme49990 were quite toasty indeed. with the output driving speakers it will of course get rid of some, but the buffers will get hotter too I would think. I wish I had one of those cheap infra-red cams