Some observation:
This is my first SS adventure. I did upgrade some caps in CD players, but never changed the circuits.
I have uploaded all my CDs to my HD. I still have about 1.2 Terra Byte free.
I had never done a CRC Pi filter with a switching PS before. Here is what I learned. The No Load Voltage, taken at the unplugged end of the PS wire plug, was 12.75 volts. Before and after my filter was installed.
The in use voltage to the Idle amp, measured at bared spots in the PS wire, was 12.68 volts.
With the CRC Pi filter inserted, The Unloaded voltage is still 12.75 volts, but, the in use voltage, measured across the final cap pair, is now 11.80 volts.
I gave up some voltage and reduced the Noise from the idled amp from obnoxious at 1:00O'clock on the Volume nob to barely able to here at 10, full scale, by listening to my speaker from within a foot. Acceptable to my direct standard 😉
The amp, at idle, draws 0.860 volts across the 7.5 ohm resistor in the CRC Pi filter.
0.860 / 7.5 = 0.11466666, call it O.1147 or 0.115 amps.
115 mA
0.115A x 11.8V = 1.357Watts
Call it 1.36 Watts at Idle.
I think this is adequate headroom, if I am calculating correctly.
I do not yet understand how the switcher is shared between the channels.
This is a new world to this old geezer. My home brews were all huge vacuum tubes. This is fun, but, I am not sure of my calculations in this environment.
I am sure of the fine sound. The LP-2020A+ rocks my world. I am not worried about the US$26.00 caps and $0.80 resistors. They were free, from my parts box 🙂
Anyone want to comment or correct my logic?
Please, feel free.
Al
This is my first SS adventure. I did upgrade some caps in CD players, but never changed the circuits.
I have uploaded all my CDs to my HD. I still have about 1.2 Terra Byte free.
I had never done a CRC Pi filter with a switching PS before. Here is what I learned. The No Load Voltage, taken at the unplugged end of the PS wire plug, was 12.75 volts. Before and after my filter was installed.
The in use voltage to the Idle amp, measured at bared spots in the PS wire, was 12.68 volts.
With the CRC Pi filter inserted, The Unloaded voltage is still 12.75 volts, but, the in use voltage, measured across the final cap pair, is now 11.80 volts.
I gave up some voltage and reduced the Noise from the idled amp from obnoxious at 1:00O'clock on the Volume nob to barely able to here at 10, full scale, by listening to my speaker from within a foot. Acceptable to my direct standard 😉
The amp, at idle, draws 0.860 volts across the 7.5 ohm resistor in the CRC Pi filter.
0.860 / 7.5 = 0.11466666, call it O.1147 or 0.115 amps.
115 mA
0.115A x 11.8V = 1.357Watts
Call it 1.36 Watts at Idle.
I think this is adequate headroom, if I am calculating correctly.
I do not yet understand how the switcher is shared between the channels.
This is a new world to this old geezer. My home brews were all huge vacuum tubes. This is fun, but, I am not sure of my calculations in this environment.
I am sure of the fine sound. The LP-2020A+ rocks my world. I am not worried about the US$26.00 caps and $0.80 resistors. They were free, from my parts box 🙂
Anyone want to comment or correct my logic?
Please, feel free.
Al
Okay, I have a quick question. One of my Lepai amps has this problem of shutting off by itself (no sound). The relay was still active (there was not "click") and the LEDs were lighted. Could this be the diodes? And if it is the diodes, what do you think causes them to shut off automatically?
Thanks in advance. 🙂
Thanks in advance. 🙂
Last edited:
I replaced the pair of 15ohm 10W with a single 8ohm 25W in my Pi filter in the stock wire from the Lepai 12V 3A switching PS to LP-2020A+ amp.
The increase of 1/2 ohm lowered the delivered voltage to the amp from 11.80V to 11.74V. This is now 1.01V less than the simple stock wire.
Both power and clarity improved. OK, so the Volume knob goes higher and I am easily fooled. At any rate, no negative effect yet noticed. What a great hobby 🙂
The 8ohm 25W drops 0.933V. So: 0.933/8 = 0.117V per one ohm.
The 7.5 ohm dropped 0.800V. So: 0.800/7.5 = 0.107V per one ohm.
Hmmm...
I need to buy a new 2020B chip and get the datasheet 🙂
The increase of 1/2 ohm lowered the delivered voltage to the amp from 11.80V to 11.74V. This is now 1.01V less than the simple stock wire.
Both power and clarity improved. OK, so the Volume knob goes higher and I am easily fooled. At any rate, no negative effect yet noticed. What a great hobby 🙂
The 8ohm 25W drops 0.933V. So: 0.933/8 = 0.117V per one ohm.
The 7.5 ohm dropped 0.800V. So: 0.800/7.5 = 0.107V per one ohm.
Hmmm...
I need to buy a new 2020B chip and get the datasheet 🙂
Upon further measuring and playing with Duncan's PSUD II numbers:
The 8 Ohm is theoretically 8.051 ohms, giving 0.933 volts.
The 7.5 ohm is theoretically 7.490 ohms, giving 0.800 volts.
So, if the modeling is accurate, the 0.933/8.051 = 0.1158V per one ohm.
So, if the modeling is accurate, the 0.800/7.490 = 0.1068V per one ohm.
I gained 0.009V with approx. 0.7ohms. Digital is so picky, picky, picky. Me, too 😉
With the WMC volume set to 100 (MAX) and playing the Stereophile Test CD -20dB pink noise signal (a minute in phase and a minute out of phase) on repeat:
Setting the LP2020A1+ Volume knob dot to just below auto-cutoff with the ~7.5 ohm was to approx. 9:30.
Now:
Setting the dot with the ~8.1 ohm to approx. 11:30 works great. I can listen to the CDs very well at 15 in the WMC, 20 if the recording has VERY quiet passages.
I have not yet even opened the LP-2020A+ 😉
More to come. What a great hobby!
The 8 Ohm is theoretically 8.051 ohms, giving 0.933 volts.
The 7.5 ohm is theoretically 7.490 ohms, giving 0.800 volts.
So, if the modeling is accurate, the 0.933/8.051 = 0.1158V per one ohm.
So, if the modeling is accurate, the 0.800/7.490 = 0.1068V per one ohm.
I gained 0.009V with approx. 0.7ohms. Digital is so picky, picky, picky. Me, too 😉
With the WMC volume set to 100 (MAX) and playing the Stereophile Test CD -20dB pink noise signal (a minute in phase and a minute out of phase) on repeat:
Setting the LP2020A1+ Volume knob dot to just below auto-cutoff with the ~7.5 ohm was to approx. 9:30.
Now:
Setting the dot with the ~8.1 ohm to approx. 11:30 works great. I can listen to the CDs very well at 15 in the WMC, 20 if the recording has VERY quiet passages.
I have not yet even opened the LP-2020A+ 😉
More to come. What a great hobby!
That pink noise is on: Stereophile Test CD Volume One, track four, for those of you that do not have them memorized 😉
Oh, the new watts numbers:
New One Ohm = 0.1158 Volts
Call it Volts one Ohm = 0.116 Amps
116 mA * 11.64 V = 1.35 Watts.
Old One Ohm = 0.1068 Volts
Call it 0.107 Volts = 0.107 Amps
107 mA * 11.8 V = 1.26 Watts
I only gained 0.09 Watts, by lowering the Voltage.
But it is much better behaved.
Digital it picky, picky, picky.
Me, too 🙂
New One Ohm = 0.1158 Volts
Call it Volts one Ohm = 0.116 Amps
116 mA * 11.64 V = 1.35 Watts.
Old One Ohm = 0.1068 Volts
Call it 0.107 Volts = 0.107 Amps
107 mA * 11.8 V = 1.26 Watts
I only gained 0.09 Watts, by lowering the Voltage.
But it is much better behaved.
Digital it picky, picky, picky.
Me, too 🙂
I managed to solve my own problem after fiddling with the Lepai in the last couple of days. This is what happened...
For some unknown reason, the amp shut off automatically after a while of operation. What I did was instead of switching off the amp using the front toggle switch, I power down from the mains. This allows the power capacitor to discharge gradually. When I switch on the Lepai next time, this capacitor is "empty" and the voltage doesn't surge up to shut down the amp. I am been switching off my Lepai this way since Tuesday and so far so good.
Maybe somebody out there can give me better advice. By the way, I am using a 5,600uf power capacitor.
Last edited:
The switching power supply has an internal, high ohm, bleeder resistor to bleed its capacitors, when it is unplugged. When the system is On, and you turn off the rocker switch on the front of the amp, the power supply just stays on. SEE: the power LED, green (on my PS wall wart).
With the amp on, the voltage into the amp on my added PS CRC Pi filter is 11.74 volts. When I turn off the amp, the voltage rises to the 12.75 volts it was before my filter. My filter is not functioning, without a load. If I then leave my multimeter attached to the Amps power plug, and unplug the switching PS, the voltage slowly drops at the plug, like ten volts in thirty seconds (or more...)
My CRC Pi filter is 20,006.8uF 😉
If you leave the amp rocker switch on and unplug the switching PS, the amp power cap bleeds along with the caps in the switcher.
Thus, when you plug it back in, the PS caps have to recharge, too. The amp PS filter cap does not get any power until the switcher is functioning.
Mystery solved?
Happy Ears!
Al
With the amp on, the voltage into the amp on my added PS CRC Pi filter is 11.74 volts. When I turn off the amp, the voltage rises to the 12.75 volts it was before my filter. My filter is not functioning, without a load. If I then leave my multimeter attached to the Amps power plug, and unplug the switching PS, the voltage slowly drops at the plug, like ten volts in thirty seconds (or more...)
My CRC Pi filter is 20,006.8uF 😉
If you leave the amp rocker switch on and unplug the switching PS, the amp power cap bleeds along with the caps in the switcher.
Thus, when you plug it back in, the PS caps have to recharge, too. The amp PS filter cap does not get any power until the switcher is functioning.
Mystery solved?
Happy Ears!
Al
OK, my final filter capacitance is ~5K uF. Each of the two ~10K uf 50 Vdc (20%) with an ESR (Equivalent Series Resistance) of 33 mOhm. their Ripple Current is 4 Amps)They are Panasonic TS-HA (the short and fat ones) They are in series, made into a Pi filter with an ~8 ohm 25W resistor between the two caps + connectors.
The final bypass capacitor the amp is an Obbligato +6.8 uF 630V PP/oil. I do not know its ESR, but it makes what the amp "sees" less than 33 mOhm.
______
"Same" caps in series double their ripple current rating and voltage rating, but halve their capacitance.
The parallel cap is not to increase the capacitance. 6.8 to 10K (20%) is not there for more capacitance but to smooth the "voice" (NOT specifified any where but hear, hear?)
The caps and resistor are not selected to add to the power from the switching power supply. They are there to smooth the Direct Current the amp "sees". They are wretched overkill. 100 volts and 8 amps and 5K uf (20%) for an circuit that is ~12 volts and 1.37 amps) You know, barely adequate for Home Brew!
Having the CRC Pi filter spliced into the wire twixt switching Power Supply and Amp means the caps stay at runtime level 24/7. That is a Good Thing 🙂 They do not draw any current without the amp turned on 😉
Sorry if this got wordy, but I wanted to tell where I was coming from and where I am going...
To Listen 8*D
Now Playing: les animaux modelas - Vernon Handley-London Symphany... again... WOW!
Happy ears!
Al
The final bypass capacitor the amp is an Obbligato +6.8 uF 630V PP/oil. I do not know its ESR, but it makes what the amp "sees" less than 33 mOhm.
______
"Same" caps in series double their ripple current rating and voltage rating, but halve their capacitance.
The parallel cap is not to increase the capacitance. 6.8 to 10K (20%) is not there for more capacitance but to smooth the "voice" (NOT specifified any where but hear, hear?)
The caps and resistor are not selected to add to the power from the switching power supply. They are there to smooth the Direct Current the amp "sees". They are wretched overkill. 100 volts and 8 amps and 5K uf (20%) for an circuit that is ~12 volts and 1.37 amps) You know, barely adequate for Home Brew!
Having the CRC Pi filter spliced into the wire twixt switching Power Supply and Amp means the caps stay at runtime level 24/7. That is a Good Thing 🙂 They do not draw any current without the amp turned on 😉
Sorry if this got wordy, but I wanted to tell where I was coming from and where I am going...
To Listen 8*D
Now Playing: les animaux modelas - Vernon Handley-London Symphany... again... WOW!
Happy ears!
Al
Great yer enjoying it.
Be aware that these are known for dying early on and unexpectedly..well not that unexpectedly given their pricings, truly poor circuit design and the horrid ...Fake.. Tripath chip imposter .
Be aware that these are known for dying early on and unexpectedly..well not that unexpectedly given their pricings, truly poor circuit design and the horrid ...Fake.. Tripath chip imposter .
Y'all are nothing if not persistent.
I haven't noticed numerous complaints of premature death among LP-2020's*, and I have no evidence that the Tripath is fake.
You wrote, "Each... with an ESR (Equivalent Series Resistance) of 33 mOhm," so it was unclear if that was taken into account.
* This doesn't include user error of course!
I haven't noticed numerous complaints of premature death among LP-2020's*, and I have no evidence that the Tripath is fake.
And ESR is doubled.
You wrote, "Each... with an ESR (Equivalent Series Resistance) of 33 mOhm," so it was unclear if that was taken into account.
* This doesn't include user error of course!
Bare:
All parts are "Fake". Name brands are just marketing drivel. The parts do what each part does, or doesn't do. I was in quality control, in the early years. Good manufacturing produces units which are exactly half over and half under. On the dot is just a design estimate. Manufacturing a million things does not deliver a million things. Some are just dead on arrival. Some squeak through QC but do not survive forever in real use. Even Rolls Royce delivers an occasional Lemon. They simply eat it, they enjoy a bit of profit margin 😉
TA2020 chips are US$3.99, onesies, free delivery. Their QC may not be perfect, but the parts seem to work OK. The CRC Pi filter in the final power supply reduces the Stress on the chip, not just my ears 😉
The 12V 3A switching PS is manufactured, too, and the CRC Pi filter allows it to wander a bit more, over time.
There was an old story about a horse buggy the was so well built, the entire thing failed, all parts, at the same time. It was just a fable. Stuff happens whenever it happens. We can just try and not get killed by the jet falling out of the sky. Some do not succeed. We trudge on.
Were the "fake" chips properly "burned in" by several hundred hours of continuous use in the real environment?
We only expect perfection, we do not all try and "coax" it to our goals.
Replacing a four buck part is not the death of the amp. It is just a four buck part. Stuff happens.
Even to hobbyists.
Sigh. I had cancer last year. Stuff happens, to every thing. Sigh.
_______
sofaspud:
Mea culpa. I blindly assumed that the CRC Pi ~8 ohm resistor swamped the first caps ESR. OK, I can live with 66 mOhm. I am fairly resigned to the fact that such rankings are not exactly "this part" measurements.
I do not even have the ESR of the big 6.8 uF PP/oil can, but I assume it is "tolerable" and its ESR is calculated with that of the two serial electrolytic caps tat it is in parallel with, even if is just the last half of them, after the resistor in parallel to determine that "seen" by the amp?
Happy Ears!
Al
All parts are "Fake". Name brands are just marketing drivel. The parts do what each part does, or doesn't do. I was in quality control, in the early years. Good manufacturing produces units which are exactly half over and half under. On the dot is just a design estimate. Manufacturing a million things does not deliver a million things. Some are just dead on arrival. Some squeak through QC but do not survive forever in real use. Even Rolls Royce delivers an occasional Lemon. They simply eat it, they enjoy a bit of profit margin 😉
TA2020 chips are US$3.99, onesies, free delivery. Their QC may not be perfect, but the parts seem to work OK. The CRC Pi filter in the final power supply reduces the Stress on the chip, not just my ears 😉
The 12V 3A switching PS is manufactured, too, and the CRC Pi filter allows it to wander a bit more, over time.
There was an old story about a horse buggy the was so well built, the entire thing failed, all parts, at the same time. It was just a fable. Stuff happens whenever it happens. We can just try and not get killed by the jet falling out of the sky. Some do not succeed. We trudge on.
Were the "fake" chips properly "burned in" by several hundred hours of continuous use in the real environment?
We only expect perfection, we do not all try and "coax" it to our goals.
Replacing a four buck part is not the death of the amp. It is just a four buck part. Stuff happens.
Even to hobbyists.
Sigh. I had cancer last year. Stuff happens, to every thing. Sigh.
_______
sofaspud:
Mea culpa. I blindly assumed that the CRC Pi ~8 ohm resistor swamped the first caps ESR. OK, I can live with 66 mOhm. I am fairly resigned to the fact that such rankings are not exactly "this part" measurements.
I do not even have the ESR of the big 6.8 uF PP/oil can, but I assume it is "tolerable" and its ESR is calculated with that of the two serial electrolytic caps tat it is in parallel with, even if is just the last half of them, after the resistor in parallel to determine that "seen" by the amp?
Happy Ears!
Al
Hi, just tried my Lepai and DC offset is bad on both channels, -0.6v on one channel and -0.8v on the other, is this a known problem? 600 and 800mV is alot, that is steady-state after a few minutes as well, it is higher to start with.
Hi,
I've been reading through this thread on and off for a few days now and thoroughly enjoyed it, even if most of it is lost on me.
I was given an LP-2020a+ recently that isn't working, like many posters before me. The Amp's LED's light-up but isn't outputting any sound from the terminals. There's also no click from the relays (I'm assuming the two yellow things on the picture I posted are the relays?)
I've tried figuring out the fault using the schematic that's on this thread, but, being new to DIY I haven't got a clue which pins are which. can anyone point me in the right direction please? Some pins are showing 12V another 10V and one at 1V. I'm using the Heatsink clip on the back of the board as Ground. Is this correct? I've also used the Ground on the power jack, too, that gives me the same reading.
Could someone perhaps mark which pins to test?
Here's the AMP. It seems to be slightly different than most in this thread.
Thanks
I've been reading through this thread on and off for a few days now and thoroughly enjoyed it, even if most of it is lost on me.
I was given an LP-2020a+ recently that isn't working, like many posters before me. The Amp's LED's light-up but isn't outputting any sound from the terminals. There's also no click from the relays (I'm assuming the two yellow things on the picture I posted are the relays?)
I've tried figuring out the fault using the schematic that's on this thread, but, being new to DIY I haven't got a clue which pins are which. can anyone point me in the right direction please? Some pins are showing 12V another 10V and one at 1V. I'm using the Heatsink clip on the back of the board as Ground. Is this correct? I've also used the Ground on the power jack, too, that gives me the same reading.
Could someone perhaps mark which pins to test?
Here's the AMP. It seems to be slightly different than most in this thread.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Thanks
NewtoDIY, if you can point me (and others 🙂) toward the circuit I'll see if I can come up with anything. Class D isn't really my thing but faultfinding still follows the same basic rules.
If you can quote the main IC type number we can hopefully pull up the data sheet that will show you how the pins are numbered etc.
If you can quote the main IC type number we can hopefully pull up the data sheet that will show you how the pins are numbered etc.
Where are the Tripath chips coming from now? Tripath went out of business years ago.
Hi, Mooly, thanks for the quick reply. You'll have to forgive my ignorance, but what does IC stand for?
- Status
- Not open for further replies.