Thanks Ed. Bummer of a TS-1 though.
Wasn't it "The Wind Crys Mary"?
Funny you should mention it, though I undertook the same thing today measuring
my speakers. I got out my little DER 5000 LCR Meter to see what it would do.
Not wanting to move the Cornwall's I figured I test them with the cables.
Got some reading on one speaker but not the other...surely not.
I moved the speakers to get to the rear of them and measured at the terminals.
did a LCR measurement of them from 100Hz, 1kHz, 10kHz, & 100kHz.
All in all both speakers measure less than 15 percent difference between
each parameter measured.
100Hz thru 100kHz
771uH.....48.19uH Ls
3.30mF....52.56nF Cs
4.50Ohm..23.34Ohm Rs
3.82Ohm..3.82Ohm (DCR)
Looking around for the loftech circuit. ? Did you post it Ed?
Are you using Victors Oscillator to keep the distortion down
when you are measuring?
Couldn't Help It.
Ahhh, if I knew how to read and follow links, I would have found it.
challenged at the moment.
Cheers,
Wasn't it "The Wind Crys Mary"?
Funny you should mention it, though I undertook the same thing today measuring
my speakers. I got out my little DER 5000 LCR Meter to see what it would do.
Not wanting to move the Cornwall's I figured I test them with the cables.
Got some reading on one speaker but not the other...surely not.
I moved the speakers to get to the rear of them and measured at the terminals.
did a LCR measurement of them from 100Hz, 1kHz, 10kHz, & 100kHz.
All in all both speakers measure less than 15 percent difference between
each parameter measured.
100Hz thru 100kHz
771uH.....48.19uH Ls
3.30mF....52.56nF Cs
4.50Ohm..23.34Ohm Rs
3.82Ohm..3.82Ohm (DCR)
Looking around for the loftech circuit. ? Did you post it Ed?
Are you using Victors Oscillator to keep the distortion down
when you are measuring?
Couldn't Help It.
Ahhh, if I knew how to read and follow links, I would have found it.
challenged at the moment.
Cheers,
Last edited:
Hey Sync.
I tried the usual through-hole dual op amp suspects the 5532, NJM2114, OPA1621 (on an adapter), OPA2134, OPA2604 etc. None have the 140 dB (DC) open loop gain and I didn't expect them to do as well. The 1621 are pretty good however.
What I haven't tried, because its in single footprint, is the AD797.
Getting quiet LME49720s is still somewhat of a challenge as many have burst noise.
I tried the usual through-hole dual op amp suspects the 5532, NJM2114, OPA1621 (on an adapter), OPA2134, OPA2604 etc. None have the 140 dB (DC) open loop gain and I didn't expect them to do as well. The 1621 are pretty good however.
What I haven't tried, because its in single footprint, is the AD797.
Getting quiet LME49720s is still somewhat of a challenge as many have burst noise.
Wayne, how would I or someone measure the burst noise of the opamps?
while I don't have that circuit that you have made, I do have a smattering
of different opamps waiting for a project (to be used).
I've got a couple of the LME 49720HAs, AD797, LT1468 (which
seems to be a great low distortion all around op amp).
A few of the OPAs etc.
Are you on an actual circuit board with yours or a proto
or bread board?
Cheers,
while I don't have that circuit that you have made, I do have a smattering
of different opamps waiting for a project (to be used).
I've got a couple of the LME 49720HAs, AD797, LT1468 (which
seems to be a great low distortion all around op amp).
A few of the OPAs etc.
Are you on an actual circuit board with yours or a proto
or bread board?
Cheers,
My oscillator is on a protoboard. I see noisy ones primarily in the notch post-filter amp I use for test.
I've observed the LME49720's burst noise in several different high-gain circuits. I think its a process/contamination/test issue.
To look for it, build a simple DC-coupled 40 dB gain non inverting amp and probe the output. Some samples have several seconds between noise bursts which is probably why they are passing through final test.
I had thought that TI's new fab for these parts (8"?) after they were brought back from EOL were clean but I'm not so sure now. My sample size is small - Vicktor has experience with this in large samples.
In noisy samples you'll likely see waveforms with very sharp steps. The ones that are good are very good.
LME49720 burst noise.
The above image was an LME49720 in an instrumentation amp on a PC board.
See also: LM4562 / LME49720 low frequency noise - Audio Amplifiers Forum - Audio Amplifiers - TI E2E Community
I've observed the LME49720's burst noise in several different high-gain circuits. I think its a process/contamination/test issue.
To look for it, build a simple DC-coupled 40 dB gain non inverting amp and probe the output. Some samples have several seconds between noise bursts which is probably why they are passing through final test.
I had thought that TI's new fab for these parts (8"?) after they were brought back from EOL were clean but I'm not so sure now. My sample size is small - Vicktor has experience with this in large samples.
In noisy samples you'll likely see waveforms with very sharp steps. The ones that are good are very good.
LME49720 burst noise.
The above image was an LME49720 in an instrumentation amp on a PC board.
See also: LM4562 / LME49720 low frequency noise - Audio Amplifiers Forum - Audio Amplifiers - TI E2E Community
Are you sure that the noise is coming from the op-amp?
Is the circuit shielded against RF?
The bursts look a bit like the kind of noise you would get from a bursted (e.g TDMA) RF noise coupling into a circuit. Some RF systems will give a periodic noise, other systems may give only a few pulses.
Is the circuit shielded against RF?
The bursts look a bit like the kind of noise you would get from a bursted (e.g TDMA) RF noise coupling into a circuit. Some RF systems will give a periodic noise, other systems may give only a few pulses.
I can confirm that the burst comes from the opamps. I saw this within years. Shielding can't help. Now around 50...60% of the LME49720/LM4562 in SO8 package have this type of noise. The very bad situation. No alternatives.
Victor.
I can replace them other type dual op amps and there is no burst noise. Vicktor has documented the same issue with the LME49720. It's somewhat lot-dependent according to Vicktor.
Its worse for Vicktor since he is using SMT. I'm lucky all I have to do is pop out a DIP.
I had some NOS LME49860 (44V LME49720) that I recently tested. All 10 were noisy.
With some of the LME49720 you have to listen for several seconds before they start to have "fits."
Its worse for Vicktor since he is using SMT. I'm lucky all I have to do is pop out a DIP.
I had some NOS LME49860 (44V LME49720) that I recently tested. All 10 were noisy.
With some of the LME49720 you have to listen for several seconds before they start to have "fits."
Last edited:
Vicktor have you ever tried the AD797? I realize that they may be too expensive.
The published open loop gain for the AD797 is 100 dB at 1 kHz; the LME49720's typical looks to be about 6 dB less. I have some AD797 to try in the VCA-stabilized oscillator.
The OPA1612 has similar 1 kHz gain but the output stage doesn't seem to have the drive capability. The part has an odd open loop Zo vs. frequency curve where it rises below 100 kHz. In the oscillator it has higher THD than the LME49720. But the OPA1612 seems to be quiet.
The published open loop gain for the AD797 is 100 dB at 1 kHz; the LME49720's typical looks to be about 6 dB less. I have some AD797 to try in the VCA-stabilized oscillator.
The OPA1612 has similar 1 kHz gain but the output stage doesn't seem to have the drive capability. The part has an odd open loop Zo vs. frequency curve where it rises below 100 kHz. In the oscillator it has higher THD than the LME49720. But the OPA1612 seems to be quiet.
The burst noise I showed earlier, aka popcorn noise, has a virtually identical signature to this example provided by ADI.
"DC Amplifier Noise Revisited," Al Ryan and Tim Scranton, Analog Dialog 18-1, 1984.
"DC Amplifier Noise Revisited," Al Ryan and Tim Scranton, Analog Dialog 18-1, 1984.
DC Amplifier Noise Revisited, Analog Dialog 18-1, 1984. - Pro Audio Design Forum
"DC Amplifier Noise Revisited," Al Ryan and Tim Scranton, Analog Dialog 18-1, 1984.
"DC Amplifier Noise Revisited," Al Ryan and Tim Scranton, Analog Dialog 18-1, 1984.
DC Amplifier Noise Revisited, Analog Dialog 18-1, 1984. - Pro Audio Design Forum
Last edited:
I had little more noise and more distortions when I tried the OPA1612 at 10kHz.
Not tried AD797. Expensive and single. It maybe has for some dB higher open loop gain, but also has the lower input impedance and probably higher input current. When it runs in inverted mode where the feedback resistors can be compared with the AD797 input impedance, the real open loop gain can drops down for the same some decibels.
That is unfortunate. There may be a correlation between the noise and leakage. I know leakage was a key issue in the testing and sorting.
You can rescale all the impedances to work with an AD797 class opamp. What you need to know is in the datasheets. it won't be a simple swap. And once optimized you can't switch back for the same reasons. One important aspect is the operating level. it seems you want it as high as possible before you run into non-linearities. The 5534 sweet-spot is around 3V. However 7 volts is possible with the newer opamps. They may need a buffer for 600 Ohms or less. The LME 49600 that's in the RTX is a really good candidate with lots of drive.
You can rescale all the impedances to work with an AD797 class opamp. What you need to know is in the datasheets. it won't be a simple swap. And once optimized you can't switch back for the same reasons. One important aspect is the operating level. it seems you want it as high as possible before you run into non-linearities. The 5534 sweet-spot is around 3V. However 7 volts is possible with the newer opamps. They may need a buffer for 600 Ohms or less. The LME 49600 that's in the RTX is a really good candidate with lots of drive.
looks like the ADA4898 comes in dual soic - don't know how 65 MHz op amps like DIP adapters
current noise about the same, lower Z would take advantage of the lower Vnoise
not audio speced but - 116 dB @ 100 kHz seems promising
only 103 dB DC gain, but linearized input stage and beefy ouput drive may make enough difference the audio distortion performance
current noise about the same, lower Z would take advantage of the lower Vnoise
not audio speced but - 116 dB @ 100 kHz seems promising
only 103 dB DC gain, but linearized input stage and beefy ouput drive may make enough difference the audio distortion performance
I think you may be on to something.
I discovered that the LME49720 (at least in this circuit) seems to have a problem with EMI.
I can set it off with a nearby cell phone while receiving data and the swept response with a grid dip meter shows a broad response to RF in the form of DC offset shift.
No other dual op amp I've tried under identical conditions shows this tendency including the OPA1612 which has similar bandwidth.
Shielding helps but its odd that no other dual op amp does this.
Its not just the Protoboard - I have circuits on PC boards where the LME49720 has burst noise and other op amps don't.
A high gain LME49720 circuit makes a good WiFi sniffer.
LME49720 Oscillator Notch Out with a Cell Phone 3" from the DUT.
OPA1612 Oscillator Notch Out with a Cell Phone 3" from the DUT.
Last edited: