HI KLe,
They are much less expensive if you get them from the source...
http://www.audio-gd.com/Pro/diy/OP/op.htm
They are much less expensive if you get them from the source...
http://www.audio-gd.com/Pro/diy/OP/op.htm
t. said:
I`m looking into improving the regulators on the board. Keeping it on board has advantages as well. No wiring to an external supply (less series inductance and less noise pick-up) and no danger of adding ground loops.
I had the following plan: there are two 1.2k resistors that go from the +-60V supply rails to the regulator circuit. These resistors can be used as part of an RC filter. My plan was to add elcaps of about 22-47uF or so from those R`s to GND to make an RC filter. That RC filter would filter out most if not virtually all rubbish that is on the power rails, not only that, it does more things, it stabilizes the voltage over the regulating transistors and current variations (opamp supply current variations induced by the audio signal that is being amplified) in those 1.2k resistors would be reduced. Stabilizing that voltage would make the current flow through the zener diodes also much more stable.
I also wanted to place an elcap at the emitter of the regulating transistor, so at the regulated voltage, this would lower the output impedance of the regulator for high frequencies so the power rails going to the opamp should be cleaner, maybe something like 22uf at that point is also nice. Did some simulations with switchcad, adding those caps, makes a big difference according to the simulations, noise on the main power rails has now far less influence on the noise on the regulated low voltage rails.
Plan to try it out on the actual modules, also plan to do some measurements. Idea for measurements: drive the modules with some input signal, the input signal should induce a varying current drawn by the opamps, this should induce voltage variations on the power supply lines going to the opamp, it should also cause variations in the voltage drop over the 1.2k resistors. Then adding a C between the 1.2k resistors and GND should drastically reduce the voltage variations over those 1.2k resistors and should also lower the voltage variations on the supply lines going to the opamp. Adding another elcap on the power lines going to the opamp should further reduce voltage variations on those lines.
Best regards
Gertjan
ghemink said:
Sounds good Gertjan and thanks for sharing.
What caps are you planning on using?
It would be nice to get some improvements on the UCD pcb first and then maybe later see how separate regs would compare
mx-5 said:
My Chinese isnt so good. Even with Babelfish translating their page I only get one contact phone number.
Purchases and delivers goods the telephone : 0757-22381191 Miss Zhong
Anyone have success ordering these from China?
Hi Daveis,
I found the link on the Burson thread... http://www.diyaudio.com/forums/showthread.php?s=&threadid=79185&perpage=10&pagenumber=3
read post #27, from Finneybear, onward! Apparently the dealer gets his english e-mails translated, so communication is slow.
I found the link on the Burson thread... http://www.diyaudio.com/forums/showthread.php?s=&threadid=79185&perpage=10&pagenumber=3
read post #27, from Finneybear, onward! Apparently the dealer gets his english e-mails translated, so communication is slow.
t. said:
For the caps on the regulated rails, probably something like a BG, maybe need to remove the ceramic SMD caps to avoid ringing. The caps that will be connected to the 1.2k resistors can be lower grade, see what I can find in the spare part box🙂
Gertjan
lm4562 current source loading
Any guesses on what side to put the current source load on anyone?
Mike
Any guesses on what side to put the current source load on anyone?
Mike
Gertjan, or anyone.
What is the regulator circuit schematic.
I can't remember. Isn't it a zener biased follower?
Mike
What is the regulator circuit schematic.
I can't remember. Isn't it a zener biased follower?
Mike
Portlandmike said:
Yes, it is a zener based emitter follower using a power darlington transistor, the zener has a 22uF elcap placed in parallel. It is a very simple circuit, the 1.2k resistor I was talking about goes from the 60V main supply rail to the collector of the darlington. From that point there is another resistor (actually two in series, at least in my UcD400 version) between the collector and base, feeding the base current and the current through the zener.
Best regards
Gertjan
Just ordered 2 pieces of 25VA/2x12V toroids and will add a regulated supply to the UcD400 AD for the low voltage. 2 standard small BRs, 2200uf/25V Panasonic FCs /7812&7912 (don't shoot) followed by a CLCRC filter (82uf.16V Panasonic FC/ 0.5mH common choke/rubycon ZA 47uf.16V/10R/Rubycon ZA 47uf/16V, followed by another Rubycon ZA already on the board.
Will let you know how it sounds as soon as I put it together, hope it will be a worthwile improvement.
Will let you know how it sounds as soon as I put it together, hope it will be a worthwile improvement.
lucpes said:
Are those regulators that noisy that you need that much filtering? I would limit myself to CLC filtering only or reduce the R value. Using a R of 10 Ohm looks a bit over the top to me. You are then also throwing away the benefits of the 0.1 Ohm output impedance that these regulators seem to have (datasheet).
Best regards
Gertjan
lucpes said:
Just finished soldering, IMO my setup got from slightly aggressive to just musical and relaxed + less noise and cleaner overall 😀 just by using the above regulated supplies for the frontend. Worthwile tweak, and you need not be an EE to build a supply. I'll attach the pdf file from Hypex which describes where to connect the wires. I omitted the resistors as per ghemink's suggestion.
The current draw is very low, you can use TL431's instead of the regs I used.
Output cap
Just read trough half of this thread
. I will continue with the rest another day. But now I have a thought about the output cap for comments:
When there is a signal on the output cap there will be forces that want to expand/contract the insulation layer inside the cap. Comments in the beginning of this thread suggest that this is a major contributor to the fact that different caps sound differently.
Based on this I have an idea to physically limit the vibration by applying pressure on the sides of the output cap. The way to do this would be two aluminum plates slightly larger than the cap with threaded holes in the corners. Small screws are used to keep and squeeze the cap between the plates.
Has anyone tried anything similar? Will it really damp vibration or will it just move resonances to higher frequencies? How about EMC? Should the aluminium plates be connected to ground?
I would appreciate your opinions before I try this mod.
Just read trough half of this thread
. I will continue with the rest another day. But now I have a thought about the output cap for comments:When there is a signal on the output cap there will be forces that want to expand/contract the insulation layer inside the cap. Comments in the beginning of this thread suggest that this is a major contributor to the fact that different caps sound differently.
Based on this I have an idea to physically limit the vibration by applying pressure on the sides of the output cap. The way to do this would be two aluminum plates slightly larger than the cap with threaded holes in the corners. Small screws are used to keep and squeeze the cap between the plates.
Has anyone tried anything similar? Will it really damp vibration or will it just move resonances to higher frequencies? How about EMC? Should the aluminium plates be connected to ground?
I would appreciate your opinions before I try this mod.
Re: Output cap
I'd try it on a cap alone first, as you might cause a short.
Might be worth a try.
I'd be more inclined to add tension via electrical means.
That is put a DC bias on them.
That would mean you'd need two caps in series of twice the desired value, then bias the common capacitor node to either positive or negative, say the supply rail, via a large omic resistor, say 1Meg.
Biggest problem with this is now you need to find a cap with twice the capacitance, and twice the voltage rating. Its already hard to find one cap that works, and is small, and has low inductance.
Best Regards,
Mike
thomsva said:Just read trough half of this thread
When there is a signal on the output cap there will be forces that want to expand/contract the insulation layer inside the cap. Comments in the beginning of this thread suggest that this is a major contributor to the fact that different caps sound differently.
Based on this I have an idea to physically limit the vibration by applying pressure on the sides of the output cap. The way to do this would be two aluminum plates slightly larger than the cap with threaded holes in the corners. Small screws are used to keep and squeeze the cap between the plates.
Has anyone tried anything similar? Will it really damp vibration or will it just move resonances to higher frequencies? How about EMC? Should the aluminium plates be connected to ground?
I would appreciate your opinions before I try this mod.
I'd try it on a cap alone first, as you might cause a short.
Might be worth a try.
I'd be more inclined to add tension via electrical means.
That is put a DC bias on them.
That would mean you'd need two caps in series of twice the desired value, then bias the common capacitor node to either positive or negative, say the supply rail, via a large omic resistor, say 1Meg.
Biggest problem with this is now you need to find a cap with twice the capacitance, and twice the voltage rating. Its already hard to find one cap that works, and is small, and has low inductance.
Best Regards,
Mike
- Status
- Not open for further replies.