Re: Great news-Power supply mods enhance the AD815's performance
I'll try it the same as you did it Ryan, I've noticed my cap values after the pre reg was much lower than what you used, I'll also try tants on the output of the last reg
Its all good fun😀
Dr.H said:
I'll try it the same as you did it Ryan, I've noticed my cap values after the pre reg was much lower than what you used, I'll also try tants on the output of the last reg
Its all good fun😀
Saw this PSU in an old magazine.. is this a bad, ok, or good PSU? 🙂
http://www.ettnet.se/~tobias/diy/psu.jpg
http://www.ettnet.se/~tobias/diy/psu.jpg
t. said:
On the AD815 and everywhere.
Btw the Jeff Rowland Coerence had an external PSU, the same size as the preamp chassis (and the same look).
With that PSU you could switch between mains, batteries (internal) and charge.
I don't know details about that PSU, but when switched to batteries it sounded even better. Not a huge difference, but noticeable.
carlosfm said:
Thats interesting Carlos, only Batteries I have is a pair of 12v 7aH Yuasa's, do you think +/- 12v would be too low? the separate supply is a good idea, I'd like to use a single case for the finished pre but at least at the minute its easy to try these things out🙂
I really wish I could find the none smd AD815, this tiny chip gets so
after a while, its a bit fiddly trying to cool this one downIt would be nice to also try a good discrete NFB regulated supplies if theres anything diy worth building
t. said:
Too low it isn't, but I always like to use a PSU voltage near the limits (between +/-15~18V).
Should be fine, though. Try it.
I have 3x 12V 4A/h batteries in series and an active rail splitter (BUF634) on my phono pre.
This gives me +/-18V. Actually a little more, but no problem.
As I use this for my phono pre, my problem now is space for more batteries around my system's rack. I have no shortage of batteries, the problem is space.
Here's some good ones, 24 A/h.
Attachments
carlosfm said:
Too low it isn't, but I always like to use a PSU voltage near the limits (between +/-15~18V).
Should be fine, though. Try it.
I have 3x 12V 4A/h batteries in series and an active rail splitter (BUF634) on my phono pre.
This gives me +/-18V. Actually a little more, but no problem.
As I use this for my phono pre, my problem now is space for more batteries around my system's rack. I have no shortage of batteries, the problem is space.
Here's some good ones, 24 A/h.
Thanks Carlos😎
I'll try them now😀 do you think I should add a small value capacitor after the batteries (as well as the one before the chip) ?
I actually almost brought a couple 24aH Yuasa's from the last radio rally I went to, if they wasn't so heavy I'd have got them but I didn't fancy carrying them around with me
carlosfm said:
You can try some bigger caps near the chip (like 470uF), bypassed.
Then snubberize.
I made this to test with batteries, but didn't test it yet.
Non SMD AD815.
Good timing Carlos😎 I've just been listening to the pre with batteries and 220uf caps, 100nf smd still on daughter board next to AD815 pins.
The sound is not bad, much more laid back than the super regs, it is nice though, the bass is not as tight
What value snubber would you recommend? the same as whats on your latest GC circuit? I presume it would be best to put the snubber on the AD815 pins if possible?
BTW nice board Carlos, looking forward to hearing your opinions with this set up
I wish I had one of those chips
t. said:
You need more capacitance, as you are not regulating.
If you had higher voltage, you could regulate.
Also, are you sure the batteries are well charged?
Anyway, try bigger caps, bypassed.
t. said:
Well, I would like to have more info on SLA batteries' output impedance.
Output impedance also increases as they discharge, and you may (audibly) notice it on these kind of circuits.
I used on this board 0.47R+1.5nF, but I'm sure it's not ideal, it may even be too far off. It's hit and miss, without any data.
I didn't search much for info yet, had no time...
If batteries have a rising impedance curve similar to most linear regs, starting very soon in the frequency, then the snubber cap must be in the uF range, not nF.
Hi all,
Rudi just to say that the preamp is working now, nothing wrong on the board at all. Just the lead i made up to test it, what a huge relief. Havent actually tested it per-say just pluggd an old internal pc speaker to the output and run a signal in from a very old (and nice actually) CD walkman. The speaker is mono so i just swapped it over 5 minutes later, and that channel worked as well.
What a bloody relief after all that graft, gonna leave it plugged int with a signal going through it for a day then test it tommorrow.
Anyway thanks for the moral support through a stressful time,
Phil
Rudi just to say that the preamp is working now, nothing wrong on the board at all. Just the lead i made up to test it, what a huge relief. Havent actually tested it per-say just pluggd an old internal pc speaker to the output and run a signal in from a very old (and nice actually) CD walkman. The speaker is mono so i just swapped it over 5 minutes later, and that channel worked as well.
What a bloody relief after all that graft, gonna leave it plugged int with a signal going through it for a day then test it tommorrow.
Anyway thanks for the moral support through a stressful time,
Phil
More is better
Different application though, on my many versions of SI Tripath amps tried batteries for a while. Ended up with 3300 ufd and 3900 ufd FM caps and 680's on board. Measured about 9500 ufd of capacitance per channel. It took this much to get the batteries where they had the slam and quickness of an active supply.
There are some fairly large FM and FC caps in 16 volts. These are perfect for a battery powered unit. Just guessing here, but 2200 ufd and 100 ufd on the power pins should make this chip sing. Add snubbers to the big caps.
I need a couple weeks to finish my current project. Afterwards another AD815 based pre is in order. Going all out on it.
George
t. said:
Different application though, on my many versions of SI Tripath amps tried batteries for a while. Ended up with 3300 ufd and 3900 ufd FM caps and 680's on board. Measured about 9500 ufd of capacitance per channel. It took this much to get the batteries where they had the slam and quickness of an active supply.
There are some fairly large FM and FC caps in 16 volts. These are perfect for a battery powered unit. Just guessing here, but 2200 ufd and 100 ufd on the power pins should make this chip sing. Add snubbers to the big caps.
I need a couple weeks to finish my current project. Afterwards another AD815 based pre is in order. Going all out on it.
George
filholder said:
Hi Phil
glad to hear that it is working. now show us your pictures.
what DC reading do you get on the output now?
I have finished the dual reg pcb design but i am just waiting to hear what these guys think sound the best before i decide on the final design
T.
I have a HUGE 225VA tranformer on my AD815 and it also have that laidback effect. when I tried a smaller 40va transformer i lost that laidback sound and it sounded a bit more stressed . i prefer it with a bigger transformer. Ryan is running a 300va on his
cheers
Rudi
The AD815 has a PSRR of -66dB!
Which probably explains why the enhanced power supply provides such an improvement in response.
T., what is the size of the toroid you're using for the AD815 power supply? I have found in both valve/tube and solid state pre's that an over-specced toriod ALWAYS improves the sound quality and bass performance.
I own a 6SN7 based pre that draws no more than 18mA and initially I was using a toroid that could supply 80mA. Yet when we increased the toriod to one rated for 5A, it made improve the performance noticeably. Same with a NAIM NAC62, which I ultimately fed with a 300VA PER CHANNEL. That pre had killer bass-resolved, deep, detailed!
Ciao
Ryan
Which probably explains why the enhanced power supply provides such an improvement in response.
T., what is the size of the toroid you're using for the AD815 power supply? I have found in both valve/tube and solid state pre's that an over-specced toriod ALWAYS improves the sound quality and bass performance.
I own a 6SN7 based pre that draws no more than 18mA and initially I was using a toroid that could supply 80mA. Yet when we increased the toriod to one rated for 5A, it made improve the performance noticeably. Same with a NAIM NAC62, which I ultimately fed with a 300VA PER CHANNEL. That pre had killer bass-resolved, deep, detailed!
Ciao
Ryan
Re: The AD815 has a PSRR of -66dB!
Right, but that's IME always the case. With more or less impressive results, but a good PSU is always an improvement for any kind of circuit.
For those who think that op-amps have very good PSRR, don't need a good PSU, etc, compare the PSRR of the AD815 with, for instance, an AD8610/20.
Negative voltage rail, frequency where it has -20db PSRR:
- AD815: ~11Mhz
- AD8610/20: ~1.5Mhz
The AD8610/20 hits 0db PSRR at ~20Mhz. 😱
The AD815 never goes below -10~-15db, and the graph on the datasheet goes up to 300Mhz.
So, the AD815 has less impressive figures at low frequencies (below 1k), but it's a very different thing at high frequencies, much better than most.
And then you wonder why the AD815 has such a clean treble... 😉
Dr.H said:
Right, but that's IME always the case. With more or less impressive results, but a good PSU is always an improvement for any kind of circuit.
For those who think that op-amps have very good PSRR, don't need a good PSU, etc, compare the PSRR of the AD815 with, for instance, an AD8610/20.
Negative voltage rail, frequency where it has -20db PSRR:
- AD815: ~11Mhz
- AD8610/20: ~1.5Mhz
The AD8610/20 hits 0db PSRR at ~20Mhz. 😱
The AD815 never goes below -10~-15db, and the graph on the datasheet goes up to 300Mhz.
So, the AD815 has less impressive figures at low frequencies (below 1k), but it's a very different thing at high frequencies, much better than most.
And then you wonder why the AD815 has such a clean treble... 😉
Sorry I'm late guys
I don't use weedy transformers for any of my projects, I've already upgraded the output transformers in my tube amp to Sowters which weigh a stone each, my 18v one for the pre is 320vA😉 I also have an 18v 120vA toroidal and a 6amp 18v E/I core
Anyway theres so much to try but it all adds to the fun.
Carlos was right, my batteries needed charging, they was dropping a fair bit on load but now seem fine🙂
I'm going to stick a couple of 470uf BC136 caps on the batteries output and try a pair of 10,000uf 16v Siemans Sikoral I have, I'll also try 0.47R+1.5nF snubbers.
I'll try other values too but to be honest I have no idea how you work out the correct values🙂
Then its back to trying the dual regs
My front room looks a mess, I've got bits and pieces everywhere and my girlfriend is not too happy🙄
I don't use weedy transformers for any of my projects, I've already upgraded the output transformers in my tube amp to Sowters which weigh a stone each, my 18v one for the pre is 320vA😉 I also have an 18v 120vA toroidal and a 6amp 18v E/I core
Anyway theres so much to try but it all adds to the fun.
Carlos was right, my batteries needed charging, they was dropping a fair bit on load but now seem fine🙂
I'm going to stick a couple of 470uf BC136 caps on the batteries output and try a pair of 10,000uf 16v Siemans Sikoral I have, I'll also try 0.47R+1.5nF snubbers.
I'll try other values too but to be honest I have no idea how you work out the correct values🙂
Then its back to trying the dual regs
My front room looks a mess, I've got bits and pieces everywhere and my girlfriend is not too happy🙄
Panelhead
Do you still have the 33ohm at the output? Or did you lower it?
I asked this because your 815 operates as a buffer and there is a different in loudness bewteen a 22ohm and 33ohms
Do you still have the 33ohm at the output? Or did you lower it?
I asked this because your 815 operates as a buffer and there is a different in loudness bewteen a 22ohm and 33ohms
33 ohms on this one
This is my third board. I have used 47 ohms, 80 ohms, and now 33. Since this resistor is outside the feedback loop, I only thought it was setting the output impedance.
My amp is 100k, was not concerned. Got some 16 ohm, will they make it a little louder?
Has anyone worked up split rail supply using LM317/337's and the LM329DZ voltage refs yet? If they really lower noise by a factor of 7, this will be Super Regs made easy.
George
jaudio said:
This is my third board. I have used 47 ohms, 80 ohms, and now 33. Since this resistor is outside the feedback loop, I only thought it was setting the output impedance.
My amp is 100k, was not concerned. Got some 16 ohm, will they make it a little louder?
Has anyone worked up split rail supply using LM317/337's and the LM329DZ voltage refs yet? If they really lower noise by a factor of 7, this will be Super Regs made easy.
George
Re: 33 ohms on this one
I tested 33 to 14ohm and you can clearly hear the differents.
Panelhead said:
I tested 33 to 14ohm and you can clearly hear the differents.
Gotta try it
I am not too worried about using the resistor as a damper, so about any value should be okay.
Were there any sound quality changes? I only used the 80 ohm becuase it was the lowest value I had with me the night I built the first buffer board. I did not like it as much as the one with gain initially. Might have been this higher value output resistor. I thought it was the ceramic caps used on the snubbers.
I have some 16 ohm, these might be better than the 33 or 22. Mine does not really need any more volume. The highest it has been turned up is maybe a hair past half rotation. This was about live music levels, 95 - 100 dB. This is why I wanted a buffer and not one with gain. The version I built with gain would never have gone to half open.
I think volume pots sound better when cranked open a bit, less series resistance.
Gotta get back to soldering.
George
I am not too worried about using the resistor as a damper, so about any value should be okay.
Were there any sound quality changes? I only used the 80 ohm becuase it was the lowest value I had with me the night I built the first buffer board. I did not like it as much as the one with gain initially. Might have been this higher value output resistor. I thought it was the ceramic caps used on the snubbers.
I have some 16 ohm, these might be better than the 33 or 22. Mine does not really need any more volume. The highest it has been turned up is maybe a hair past half rotation. This was about live music levels, 95 - 100 dB. This is why I wanted a buffer and not one with gain. The version I built with gain would never have gone to half open.
I think volume pots sound better when cranked open a bit, less series resistance.
Gotta get back to soldering.
George
- Status
- Not open for further replies.