I currently own a DIYparadise monica usb. The board has two stages, the DAC and a gain stage. It is currently powered by a 12V battery, partly because it produced a very well regulated voltage and partly because I'm just too lazy to build a proper power supply. Sounds great, but I would like to improve the sound even more.
The gain stage supposedly produces a clearer sound if powered by 18V. That gets me to my question.
What do you think of this schematic?:
I am planning on having a LM317 regulator after the voltage step up to regulated de voltage down to 18V.
Are the better alternatives?
The gain stage supposedly produces a clearer sound if powered by 18V. That gets me to my question.
What do you think of this schematic?:
I am planning on having a LM317 regulator after the voltage step up to regulated de voltage down to 18V.
Are the better alternatives?
The MAX668 is supposed to operate a 500kHz in that schematic. To me that seems pretty high.
What I am trying to know is how suitable that circuit is for high-end audio applications?
What I am trying to know is how suitable that circuit is for high-end audio applications?
500Khz is fine, you should probably use a logic level Fet (IRF7455 would be good) for that controller because it only drives with 5V. Have you used the Max668 before? I don't know if it will start up with the SHDN line connected to LDO like that, I would connect to VCC just to be safe. I have had other step down controllers from national semi. not start up because I connected the shutdown line to the internal regulator pin. Just my 2 cents.
One question: what type of output caps are you attempting to use?
500 Khz is way too high a frequency for electrolytics, even lo-Z ones. It may be even borderline for low-Z chip tantalums. see attached image from Kemet's catalog.
The problem is aggravated with the boost topology, since it delivers pulsating currents to the output.
Your best bet is to bypass them with MLCC caps.
10 uF, 25 volt X7R devices will do a wonderful job.
Otherwise, you'll have to decrease your frequency substantially.
500 Khz is way too high a frequency for electrolytics, even lo-Z ones. It may be even borderline for low-Z chip tantalums. see attached image from Kemet's catalog.
The problem is aggravated with the boost topology, since it delivers pulsating currents to the output.
Your best bet is to bypass them with MLCC caps.
10 uF, 25 volt X7R devices will do a wonderful job.
Otherwise, you'll have to decrease your frequency substantially.
Be Lazy
Get two 9volt batteries, wire em in series and carry on being lazy. If you must get the iron out, build a charger for the two lead acids.😀 😀 😀
mikee55🙂
Get two 9volt batteries, wire em in series and carry on being lazy. If you must get the iron out, build a charger for the two lead acids.😀 😀 😀
mikee55🙂
500Khz would be fine using a combination of electrolytics and ceramic in parallel. I have made many boost supplies using electrolytic and ceramic combos at this frequency and higher, with decent output ripple (<150 mV). Just be sure to "double stack" (place two ceramic caps in series with the output and ground). Ceramics have been know to fracture and cause shorts. This case is extreamly bad in a boost topology since they are connected to Vbat through the inductor and diode.
How about two LM2587s in parallel? You can sync the slave off the master thru the COMP pin, and ground the Voltage feedback pin on the slave. This makes the slave want to run "wide-open", but with each timing pulse coming thru the COMP pin from the master to the slave, it forces the slave to lock to the master's clock. Nat Semi had this circuit in their tricks book, I'll see if I can find it.... . ...... .... ..... .. ..... .. . ..... ..... ..... ........ . ... ... ...........
OK, it must be on my work CPU. As soon as I get on it I will send the ckt. Basically, you're just paralleling two 5A Step-up SimpleSwitchers running at ~100kHz- nothing too exotic.
Steve
OK, it must be on my work CPU. As soon as I get on it I will send the ckt. Basically, you're just paralleling two 5A Step-up SimpleSwitchers running at ~100kHz- nothing too exotic.
Steve
Why would you have to parallel two 5A switchers to run just a DAC and line stage?N-Channel said:
This goes back a few years, but I remember the Simple Switchers as being quite noisy.
I see, since the current draw will probably under 100 MA paralleling two 5A switchers would be way oversized.
I was looking at that schematic from National semiconductors:
http://www.national.com/pf/LM/LM2731.html
It seems that it would be modifiable for my application. (12-13VDV-18VDC Step-up converter) Would anyone know what resistor values to change?
It seems somewhat simpler that the Maxim circuit. And does run at 600kHZ/1.6MHZ, so I don't believe noise would be a problem.
I was looking at that schematic from National semiconductors:
http://www.national.com/pf/LM/LM2731.html
It seems that it would be modifiable for my application. (12-13VDV-18VDC Step-up converter) Would anyone know what resistor values to change?
It seems somewhat simpler that the Maxim circuit. And does run at 600kHZ/1.6MHZ, so I don't believe noise would be a problem.
psykar said:
I have the schematic, but I can't shrink it to below the 100kB size limit. Iwill try to find another way to get around this, or I will re-create it in Express SCH, and post the pic. Stay Tuned.......
paulb said:
Actually, they are rather quiet. As with any switched-mode regulator or power supply, SimpleSwitchers can be noisy if not filtered properly. I have made literally dozens of booster ckts using the LM2577 and '2587 to power laptops from cars, and they have ALL worked well.
As for the "Why parallel?", in looking at the schematic psykar posted, I saw the 18V @ 3.1 A out, and assumed he would need lots of current. But if for only a DAC and a linestage, even one LM2577 (3Amps) should suffice.
- Status
- Not open for further replies.