And....
TeddyPardo SuperTeddyReg
"The best regulator I've built so far" - by Teddy Pardo
I've not read past page three of the thread yet but it kind of looks like it might not do for areas where current is supplied in 'gulps' as the voltage might droop. So maybe for the OPA1632D but not the LME49600??
TeddyPardo SuperTeddyReg
"The best regulator I've built so far" - by Teddy Pardo
I've not read past page three of the thread yet but it kind of looks like it might not do for areas where current is supplied in 'gulps' as the voltage might droop. So maybe for the OPA1632D but not the LME49600??
Last edited:
SO do we just wait for a message before payment? I can pay any time, but I dont wanna mess anything up 🙂
I use the following two solders exclusively:
1. Cardas Quad Eutectic 0.032 - Cardas Solder
2. Nexxtech Standard Rosin Core 0.032
I really like that exact thickness, and both of the above have quite a bit of flux core which makes them flow well. I don't believe in any way that the Cardas stuff sounds any better, but it is by far the best solder I've ever used in terms of wetting and cleaning. I finished off my 1lb roll doing the first three boards for this run of The Wire, and now I'm faced with the unpleasant task of buying another $85 roll. It's also a reminder that I probably solder a little too much.
The best advice I can give for the surface mount stuff is this:
1. Don't use thin diameter solder - it doesn't have enough flux in the core to flow properly and make a nice joint
2. Don't use fine point tips - they have very low thermal mass, and take far too long to get the joint up to temp.
3. Have some good quality pre-fluxed solder wick on hand to deal with the inevitable solder bridges.
The worst thing people do is immediately go out and buy fine point tips and ultra-thin solder because they mistakenly think that the small parts will require them. In reality, it's almost impossible to solder with that combination of tip and solder, and you're bound to fail. Instead, you ideally want a 2-2.5mm chisel tip and a good flux core 0.032" solder.
I've actually used much larger tips, and they're quite versatile because they allow you to solder the larger TO-263 parts without changing tips. I've attached a picture below to give you an idea of the tip I used to solder all of the boards pictured in this thread. It's a massive 3.5mm tip, but it has a huge thermal mass and makes soldering the 0805 and SOIC packages a breeze.
Cheers,
Owen
1. Cardas Quad Eutectic 0.032 - Cardas Solder
2. Nexxtech Standard Rosin Core 0.032
I really like that exact thickness, and both of the above have quite a bit of flux core which makes them flow well. I don't believe in any way that the Cardas stuff sounds any better, but it is by far the best solder I've ever used in terms of wetting and cleaning. I finished off my 1lb roll doing the first three boards for this run of The Wire, and now I'm faced with the unpleasant task of buying another $85 roll. It's also a reminder that I probably solder a little too much.
The best advice I can give for the surface mount stuff is this:
1. Don't use thin diameter solder - it doesn't have enough flux in the core to flow properly and make a nice joint
2. Don't use fine point tips - they have very low thermal mass, and take far too long to get the joint up to temp.
3. Have some good quality pre-fluxed solder wick on hand to deal with the inevitable solder bridges.
The worst thing people do is immediately go out and buy fine point tips and ultra-thin solder because they mistakenly think that the small parts will require them. In reality, it's almost impossible to solder with that combination of tip and solder, and you're bound to fail. Instead, you ideally want a 2-2.5mm chisel tip and a good flux core 0.032" solder.
I've actually used much larger tips, and they're quite versatile because they allow you to solder the larger TO-263 parts without changing tips. I've attached a picture below to give you an idea of the tip I used to solder all of the boards pictured in this thread. It's a massive 3.5mm tip, but it has a huge thermal mass and makes soldering the 0805 and SOIC packages a breeze.
Cheers,
Owen
Attachments
If you want to see a very tacky video I made back when the first version of "The Wire" was being built, you can check it out here:
MVI_5833.AVI - YouTube
It has had a whopping 81 views since it went up about a year ago, so there are clearly better videos out there to show you how to solder!
Cheers,
Owen
MVI_5833.AVI - YouTube
It has had a whopping 81 views since it went up about a year ago, so there are clearly better videos out there to show you how to solder!
Cheers,
Owen
NIIICE lighting =P i'm using my Jedi forces, but i'm going to have to assume that was a good joint hehe. It covers all the right points though, so still worth watching for those who are just learning SMD. i will say though, that those just learning soldering the second side of the diode without holding it will actually be a good test of whether the iron is too hot or not, because smd diodes are probably the quickest to heat through and melt the other side, except for some low value resistors. perfectly fine if, as Owen said the iron is the right temp.
Nice to see you use cardas quad too despite the health risks; i just keep a well ventilated work area. ive tried so many others in an attempt to find an equal composition but smaller gauge for the occasional joint and the closest ive found is kester for that and Wonder leaded solder for the same size. I never did find one as good though, so i just bought another 1lb roll. i use a touch of cardas rosin flux paste too on stubborn joints. nothing to do with it being silver solder, its just GOOD solder
wolfsin: i'm thinking thats whats currently known as the minigold super regulator and it uses reasonable size power transistors for the pass transistor, so although still a bit borderline for this amp and low impedance headphones, it should be ok. i would recommend any other type pf super reg though, not enough current unless as myself and Ian have said you make the mod to supply the front end and output stage separately. really though if you are going to do that, it would be better to try and focus on the buffers as they have much lower PSRR
Last edited:
I think opc is working on a wiki post to clarify some of these requirements and I'm sure there is some posts about this earlier in the thread but I had a dig earlier in the week but haven't tracked down anything yet. So for those who've built these already, whats an indication of what is 'enough' current?
nope, regular super regulators are usually lucky to supply 100ma and most less, which isnt enough under any circumstance. the input opamps wont be using much as they are driving the buffer's input only and the buffers will depend on the headphone load, so unless opc is adding a load impedance vs supply current graph you'll still have to do some maths of your own. even then, headphones often present a different load depending on the frequency of the signal.
Notice that the TeddyReg is more like a buffered voltage reference. You have no feedback from the output but this isn't a problem if you accept voltage drops (not a constant output voltage).
Maybe that slightly shifting rail just reduces the audio signals dynamic transient a tiny bit? No other adverse effects? Maybe that results in a slightly softer sound that might suit some Hi-Fi systems? Maybe sufficient capacitance would reduce or remove that slight transient droop? As Owen stated earlier, the caps would be supplying the higher frequency transient energy.
This ones claims suggest it's pretty impressive at holding the voltage steady, but rather expensive (here in the UK anyway).
http://www.belleson.com/compareoptima.php
http://www.belleson.com/comparesuperted.php
This ones claims suggest it's pretty impressive at holding the voltage steady, but rather expensive (here in the UK anyway).
http://www.belleson.com/compareoptima.php
http://www.belleson.com/comparesuperted.php
Last edited:
96 views now 🙂
What temp do you keep the iron tip at for this work? I think mine's about 300 to 350°C. Maybe that's a bit hot.
I was surprised to hear you say you still used a sponge. For me they used to have an adverse effect of the cleanness of the tip. I use the Xytronic tip cleaner. It's brass 'wool', no water filling, no mess, clean tip. After a while the residual solder in there helps to tin the tip. I don't know how I'd solder without this tool.
XYTRONIC|460|TIP CLEANER | Farnell United Kingdom
Xytronic Solder Tip Cleaner - Rapid Electronics
Last edited:
Thanks all for the soldering tips. Definitely changing the tip, practicing on scrap boards, and working in dimly lighted areas 🙂
I think qusp nailed the supply -- it goes by many names. Glad to know it is marginal. The question about how much current IS enough was a good one.
I think qusp nailed the supply -- it goes by many names. Glad to know it is marginal. The question about how much current IS enough was a good one.
Ceteris Paribus Ratznest
I connected a beefier shunt supply, one channel with 'wire', t'uther with LME49710+49600 board, balanced for my unbalanced earz, and cannot say one izBetter than t'uther. This is to say, primarily, that an Se-Se rework board with a pair of 710's (or indeed a single 49720) might not test as well but would still sound great. YMMV
No, qusp, ratznest image will not be posted.
I connected a beefier shunt supply, one channel with 'wire', t'uther with LME49710+49600 board, balanced for my unbalanced earz, and cannot say one izBetter than t'uther. This is to say, primarily, that an Se-Se rework board with a pair of 710's (or indeed a single 49720) might not test as well but would still sound great. YMMV
No, qusp, ratznest image will not be posted.
Breakthru
You guyz have helped Wolfsin more than you can imagine. Owen's separating the power supplies on the Bal-Bal, qusp's demand for pix and his super sharp eyez in spotting the inadequate regulator have bolted Wolfsin into audio nirvana. That regulator was enough for one channel so, as I reported, it sounded great.
When I cobbled together the ratznest it was the first time the beefier powsupp had been driving only four LME49600's and I was unable to say which channel sounded better. Truth was they BOTH sounded better than I had ever heard before! Always previously the beefy guy had been driving eight LME49600s and was overburdened (and ran hot) but I could not see the obvious :-(
My interest in the LME49600 predates "the wire" but it took not only this fine design but the interaction with true fanatics to refine my implementation. Thanks all.
You guyz have helped Wolfsin more than you can imagine. Owen's separating the power supplies on the Bal-Bal, qusp's demand for pix and his super sharp eyez in spotting the inadequate regulator have bolted Wolfsin into audio nirvana. That regulator was enough for one channel so, as I reported, it sounded great.
When I cobbled together the ratznest it was the first time the beefier powsupp had been driving only four LME49600's and I was unable to say which channel sounded better. Truth was they BOTH sounded better than I had ever heard before! Always previously the beefy guy had been driving eight LME49600s and was overburdened (and ran hot) but I could not see the obvious :-(
My interest in the LME49600 predates "the wire" but it took not only this fine design but the interaction with true fanatics to refine my implementation. Thanks all.
absolutely not, super regulators are about the worst thing you would want to put large capacitance after they may even oscillate driving that much of a capacitive load.
hehe i see peranders couldnt help but pop the head in for a quick jibe at the competition lol (he makes super regs too) you really should try and resist that PA. since yours is utterly unsuitable as far as current for anything but the input stage on this amp, which has such huge PSRR its not worth worrying about imo, but doesnt have enough current for the buffers., which could if anything take advantage of the cleaner power
@ wolfsin
really glad the heavens opened up for you buddy =) glad to be of help. the minigold or coffin super regulator is an unusually high current super reg, but still not quite there, which is what i think you have an old variant of, going by the latest build mentions max current of 720ma @ +/-30v, you can tweak the current gain as below, but i reckon youde really still be pushing it even if you lowered R6 and R16 right down, increased Q3&Q4 . Q13&Q14 current and with particularly high HFE on the main transistors it would be borderline depending on Vin and does sound like thats confirmed with your testing. its otherwise quite a versatile reg as it allows much high voltages than the average one due to the power transistors.
i have a dual and a single and their pretty decent and yeah will run up to +/-80vdc from memory. by the numbers on the last version you could probably do it, but not sure whether the older ones would. heatsinking would need to be plentiful and you would need to tune for lowest voltage drop while still giving it enough to work with and not ask it to drop too much voltage on the input
from the webpage coffin gold minireg
Last edited:
I'm using Ultraclean too, but with smd soldering work I'll use some flux on pads.
Attachments
yep, thats the stuff, i use that on just about every joint. takes a good scrum with demineralized water to clean the greasy residue off, but mainly for aesthetic reasons as it doesnt do harm, being noclean
watch you dont get addicted to the smell though, its a bit sickly sweet and not good for you at all, use in a ventilated space, works a treat though
