hailteflon: Please elaborate when you have some time ... even if it means starting a new thread ... 
Teflon as an insulator of conductors = Hooray and hallaloouia.
Teflon as capacitor = under the right conditions, good enough got g'ment work. ... For audio, plastic, polystyrene, polypropylene, MKT type capacitors seem to be better, cheaper, more readily available and more than good enough for the best quality audio.
... there is considerable info on the 'Net about this.
Teflon as an insulator of conductors = Hooray and hallaloouia.
Teflon as capacitor = under the right conditions, good enough got g'ment work. ... For audio, plastic, polystyrene, polypropylene, MKT type capacitors seem to be better, cheaper, more readily available and more than good enough for the best quality audio.
... there is considerable info on the 'Net about this.Fast Eddy:
I agree with you about most everything. I know what you meant about the source of electrons.
So there are people who don't like teflon caps? I'll have to look this one up on the web. I find that teflon is far superior. Periodically there are great deals on CRC hermetically sealed teflons on ebay. CRC closed down a couple of years ago. Teflon is expensive because it takes so much time to wind the flexible dielectric. In the early 1970s CRC charged more for polycarb than for teflon. Teflon was in production in the 1950s for the military.
What I wrote for the original poster was in relation to hand-made boards using 24 ga. phone wire and solderable perf-board. They are good for about 50KHz with no sluggish behavior, which is fine for audio.
Yes, the star-ground has it's place, and must be properly managed. The use of flat circuit board tracks is a differernt situation.
When circuit builders are looking for the right op-amp they are usually looking for the op-amp that sounds best with the stray coupling on their board, unles they have everything de-coupled properly.
I have to inventory on this 'puter and am already tired of it. It is not the time so much as the fatigue. I'm trying to stay with inventory only. When I have more time I will look you up.
There is also a Signetics tutorial that is very good.
This is a good forum because it is well moderated such as to keep the roving anarchists out. Mark
I agree with you about most everything. I know what you meant about the source of electrons.
So there are people who don't like teflon caps? I'll have to look this one up on the web. I find that teflon is far superior. Periodically there are great deals on CRC hermetically sealed teflons on ebay. CRC closed down a couple of years ago. Teflon is expensive because it takes so much time to wind the flexible dielectric. In the early 1970s CRC charged more for polycarb than for teflon. Teflon was in production in the 1950s for the military.
What I wrote for the original poster was in relation to hand-made boards using 24 ga. phone wire and solderable perf-board. They are good for about 50KHz with no sluggish behavior, which is fine for audio.
Yes, the star-ground has it's place, and must be properly managed. The use of flat circuit board tracks is a differernt situation.
When circuit builders are looking for the right op-amp they are usually looking for the op-amp that sounds best with the stray coupling on their board, unles they have everything de-coupled properly.
I have to inventory on this 'puter and am already tired of it. It is not the time so much as the fatigue. I'm trying to stay with inventory only. When I have more time I will look you up.
There is also a Signetics tutorial that is very good.
This is a good forum because it is well moderated such as to keep the roving anarchists out. Mark
" ... When circuit builders are looking for the right op-amp they are usually looking for the op-amp that sounds best with the stray coupling on their board, unless they have everything de-coupled properly. ..."
Is so ... there's a lot of really good stuff gets bypassed or overlooked. There is also the "not invented here" question and the "old school" ties question to be overcome with any newer approach. There are those who insist the old '741 is still viable too ... and I still like to use mini-DIP sockets when building on perf board, even though I know that there are quality/reliability questions ...
Got a link to the Signetics' tutorial?
Is so ... there's a lot of really good stuff gets bypassed or overlooked. There is also the "not invented here" question and the "old school" ties question to be overcome with any newer approach. There are those who insist the old '741 is still viable too ... and I still like to use mini-DIP sockets when building on perf board, even though I know that there are quality/reliability questions ...
Got a link to the Signetics' tutorial?
OPAMP EXPERIENCES AND HOW TO
Instead of sticking a 10 ohm resistor in the supply rails.
As one example, IMHO, you are far better off sticking VREGs such as the 78L12 and 79L12 there instead, f you have +-15V supply rails. 47-100uF with 100n poly.in parallel near the supply pins is then all that is needed.
You then end up with a more neutral sound and no stability problems , even with the faster opamps where the manufacturers
recommend 100nF ceramic or similar. Using ceramic capacitors in these locations often results in a more aggressive sounding top end. Using this method with an AD8066 gives me excellent results, and the sound is more evenly balanced.
SandyK
"The proof of the pudding, is in the eating"
Instead of sticking a 10 ohm resistor in the supply rails.
As one example, IMHO, you are far better off sticking VREGs such as the 78L12 and 79L12 there instead, f you have +-15V supply rails. 47-100uF with 100n poly.in parallel near the supply pins is then all that is needed.
You then end up with a more neutral sound and no stability problems , even with the faster opamps where the manufacturers
recommend 100nF ceramic or similar. Using ceramic capacitors in these locations often results in a more aggressive sounding top end. Using this method with an AD8066 gives me excellent results, and the sound is more evenly balanced.
SandyK
"The proof of the pudding, is in the eating"
The 10 ohm resistors don't go into rails, they go to the IC terminals.
go to: datasheetarchive.com
Search for AN142, AN165, AN166. Signetics went to Phillips and Phillips/NXP doesn't seem to want us to read these anymore, but DSA has them.
DSA has a daily limit of about 20 downloads. Disconnect from them to reset the limit after you have gotten about 10 datasheets.
go to: datasheetarchive.com
Search for AN142, AN165, AN166. Signetics went to Phillips and Phillips/NXP doesn't seem to want us to read these anymore, but DSA has them.
DSA has a daily limit of about 20 downloads. Disconnect from them to reset the limit after you have gotten about 10 datasheets.
OP AMP EXPERIENCES AND HOW TO
hailteflon
I presume you meant non inverting and inverting pins, and NOT supply pins ? If so, this is an old technique that was used on devices like the LM379 dual power amplifier.
Yes , the -ve supply pin is regarded as another input in devices like the AD8066 etc, and needs close bypassing.
I still believe that dual regulators close to the supply pins is a better option. Not only does it further reduce supply ripple, but the lower impedance in conjuction with the often used 100uF/100nF combination helps achieve a more neutral tonal balance.Remember that Philips/Marantz is the same mob that decided to place a series RC network in their earlier model CD player's supply lines to further reduce ripple, and opened up a new can of worms, in the way of supply modulation between the dual amplifiers etc. (CD65 etc)
I don't exactly have a great deal of faith in their earliest application notes after this.
SandyK
hailteflon
I presume you meant non inverting and inverting pins, and NOT supply pins ? If so, this is an old technique that was used on devices like the LM379 dual power amplifier.
Yes , the -ve supply pin is regarded as another input in devices like the AD8066 etc, and needs close bypassing.
I still believe that dual regulators close to the supply pins is a better option. Not only does it further reduce supply ripple, but the lower impedance in conjuction with the often used 100uF/100nF combination helps achieve a more neutral tonal balance.Remember that Philips/Marantz is the same mob that decided to place a series RC network in their earlier model CD player's supply lines to further reduce ripple, and opened up a new can of worms, in the way of supply modulation between the dual amplifiers etc. (CD65 etc)
I don't exactly have a great deal of faith in their earliest application notes after this.
SandyK
SandyK: No, the filters are on the supply pins. Ten ohms or less will not degrade the signal. These filters are not for ripple, they are for hi-freq noise.
Regulators eat about 3 volts. Op-amps seem to work better near max supply. Spec sheets specify this.
Turns out that Phillips/NXP does supply the AN 142, 165, 166 on their site and also still makes the NE5532. Had to look for them a little while.
Regulators are a bit noisy, as I recall. They have a -3db bandwidth in the specs, so they are probably some sort of simple differential amp.
Regulators eat about 3 volts. Op-amps seem to work better near max supply. Spec sheets specify this.
Turns out that Phillips/NXP does supply the AN 142, 165, 166 on their site and also still makes the NE5532. Had to look for them a little while.
Regulators are a bit noisy, as I recall. They have a -3db bandwidth in the specs, so they are probably some sort of simple differential amp.
OP AMP EXPERIENCES AND HOW TO
Hailteflon
The main place I use the technique mentioned, is in a modified Musical Fidelity X-DAC V3 . I have replaced the very ordinary NE5532s with OPA2134s, and I am using an AD8066 in the output stage. The X-DAC normally runs on supplies of around
+-15V (unregulated) I am using a PSU with +-15V regulated output for vastly improved sonic performance.
The AD8066 requires no more than +-12V , so this way I get exactly what I need , with no performance penalties to the AD8066. The OPA2134s seem to perform very well on +-12V too.
SandyK
Hailteflon
The main place I use the technique mentioned, is in a modified Musical Fidelity X-DAC V3 . I have replaced the very ordinary NE5532s with OPA2134s, and I am using an AD8066 in the output stage. The X-DAC normally runs on supplies of around
+-15V (unregulated) I am using a PSU with +-15V regulated output for vastly improved sonic performance.
The AD8066 requires no more than +-12V , so this way I get exactly what I need , with no performance penalties to the AD8066. The OPA2134s seem to perform very well on +-12V too.
SandyK
This is turning into a very informative discourse ...
SandyK: " ... 47-100uF [on the PS rails] with 100n poly. [or ceramic] in parallel near the [op-amp] supply pins ... with the faster opamps where the manufacturers recommend 100nF ceramic or similar. Using ceramic capacitors in these locations often results in a more aggressive sounding top end. Using this method with an AD8066 gives me excellent results ... "
Clarifications: "these locations" being as close as possible to the op-amp power pins, thusly?:
For the Analog AD8085 & AD8066 = http://www.analog.com/UploadedFiles/Data_Sheets/AD8065_8066.pdf = see schematics on page 15, etc. ... as shown, the ceramic or plastic / poly caps (the 100nF "snubbing caps"), would be closely connected to pins Vs(-) #4 and Vs(+) #8 and, if possible, the ceramic or plastic / poly caps should also be connected very close to the local signal and/or split rail power ground ("close coupled" to power pins and power ground). ... This is the "performance secret" many of these manufacturers use to get those really great specs in their data sheets ...
I have used the plastic, yellow MKT type caps extensively for the "snubbing" caps with great success, IMHO ...
(Note also that this Analog AD8066 8-pin configuration shares a common, dual channel op-amp pin configuration with several manufacturers and many dual op-amp substitutions are possible with similarly improved in the results.)
SandyK: " ... 47-100uF [on the PS rails] with 100n poly. [or ceramic] in parallel near the [op-amp] supply pins ... with the faster opamps where the manufacturers recommend 100nF ceramic or similar. Using ceramic capacitors in these locations often results in a more aggressive sounding top end. Using this method with an AD8066 gives me excellent results ... "
Clarifications: "these locations" being as close as possible to the op-amp power pins, thusly?:
For the Analog AD8085 & AD8066 = http://www.analog.com/UploadedFiles/Data_Sheets/AD8065_8066.pdf = see schematics on page 15, etc. ... as shown, the ceramic or plastic / poly caps (the 100nF "snubbing caps"), would be closely connected to pins Vs(-) #4 and Vs(+) #8 and, if possible, the ceramic or plastic / poly caps should also be connected very close to the local signal and/or split rail power ground ("close coupled" to power pins and power ground). ... This is the "performance secret" many of these manufacturers use to get those really great specs in their data sheets ...
I have used the plastic, yellow MKT type caps extensively for the "snubbing" caps with great success, IMHO ...
(Note also that this Analog AD8066 8-pin configuration shares a common, dual channel op-amp pin configuration with several manufacturers and many dual op-amp substitutions are possible with similarly improved in the results.)
I did notice a very slight "mistake" made by the Analog folks on a schematic diagram on page 17 of their AD8065 / AD8066 app notes:
Figure 52 : Single Supply. IMHO, I would have added more capacitance, using a 10 uF electrolytic and 150 nF or 200 nF plastic or ceramic cap here. More is better and this increase would "balance" the PS rails a little better compared to the other split rail supply diagrams. ... I know: picky, picky, picky ...
Also, take a look at Figure 60 on page 24 ( http://www.analog.com/UploadedFiles/Data_Sheets/AD8065_8066.pdf ) ... A well thought out Instrumentation Amplifier = this is an absolutely first class Balanced to Unbalanced audio line / cable adapter circuit. Of interest is that the Analog engineers used two, dual channel AD8066 chips to get this done, the two input op-amps being on separate chips with doubly redundant PS rail filter cap sets ... in order to get maximum power supply noise rejection (maximum PSRR) and maximum separation (maximum CMRR) ... This is also a very good summing amp circuit = two channels in and one channel out for a sub woofer driver, a low pass filter circuit can be placed between the input pair and the output op-amp ... I'm building on of these sub woofer drivers on my kitchen table as I write. (The extra op-amp channel can be used to drive some blinky, blinky LEDs as well) ...
Figure 52 : Single Supply. IMHO, I would have added more capacitance, using a 10 uF electrolytic and 150 nF or 200 nF plastic or ceramic cap here. More is better and this increase would "balance" the PS rails a little better compared to the other split rail supply diagrams. ... I know: picky, picky, picky ...
Also, take a look at Figure 60 on page 24 ( http://www.analog.com/UploadedFiles/Data_Sheets/AD8065_8066.pdf ) ... A well thought out Instrumentation Amplifier = this is an absolutely first class Balanced to Unbalanced audio line / cable adapter circuit. Of interest is that the Analog engineers used two, dual channel AD8066 chips to get this done, the two input op-amps being on separate chips with doubly redundant PS rail filter cap sets ... in order to get maximum power supply noise rejection (maximum PSRR) and maximum separation (maximum CMRR) ... This is also a very good summing amp circuit = two channels in and one channel out for a sub woofer driver, a low pass filter circuit can be placed between the input pair and the output op-amp ... I'm building on of these sub woofer drivers on my kitchen table as I write. (The extra op-amp channel can be used to drive some blinky, blinky LEDs as well) ...
I wish I had the time to get deeper into this conversation.
SandyK: I looked at the BurrB OPA2134 data and found the following.
"The p-channel JFETs in the
FET input stage exhibit a varying input capacitance with
applied common-mode input voltage. In inverting configurations
the input does not vary with input voltage since the
inverting input is held at virtual ground. However, in
noninverting applications the inputs do vary, and the gateto-
source voltage is not constant. The effect is increased
distortion due to the varying capacitance for unmatched
source impedances greater than 2kW.
To maintain low distortion, match unbalanced source impedance...."
JFET front ends have their idiosyncracies.
I like the NE55xx series because there is zero input-output phase lag at a gain of Av=40 from 20 to 20KHz. The OPA37 has the same property, which I feel is a great advantage in audio. The OPA2134 data does not give gain-phase response except for open-loop. Perhaps a little research on bode plots would reveal how to interpret closed-loop phase response from open-loop graph.
Nordic: The resistor and the capacitors ( 1uf tant and .1 uf film) at the PS pin form a low-pass filter. I would think twice before putting an inductor there.
SandyK: I looked at the BurrB OPA2134 data and found the following.
"The p-channel JFETs in the
FET input stage exhibit a varying input capacitance with
applied common-mode input voltage. In inverting configurations
the input does not vary with input voltage since the
inverting input is held at virtual ground. However, in
noninverting applications the inputs do vary, and the gateto-
source voltage is not constant. The effect is increased
distortion due to the varying capacitance for unmatched
source impedances greater than 2kW.
To maintain low distortion, match unbalanced source impedance...."
JFET front ends have their idiosyncracies.
I like the NE55xx series because there is zero input-output phase lag at a gain of Av=40 from 20 to 20KHz. The OPA37 has the same property, which I feel is a great advantage in audio. The OPA2134 data does not give gain-phase response except for open-loop. Perhaps a little research on bode plots would reveal how to interpret closed-loop phase response from open-loop graph.
Nordic: The resistor and the capacitors ( 1uf tant and .1 uf film) at the PS pin form a low-pass filter. I would think twice before putting an inductor there.
OP AMP EXPERIENCES AND HOW TO
hailteflon
The Silicon Chip magazine design "Studio Series Stereo Preamplifier' uses an OPA2134 preceded by a volume control as a non inverting buffer. In their design they made provision on the PCB to fit a resistor between output and inverting input. The value of the resistor was chosen to be the same as the resistance from volume control wiper to earthy end at most used
listening setting . This resistor was optional.
SandyK
hailteflon
The Silicon Chip magazine design "Studio Series Stereo Preamplifier' uses an OPA2134 preceded by a volume control as a non inverting buffer. In their design they made provision on the PCB to fit a resistor between output and inverting input. The value of the resistor was chosen to be the same as the resistance from volume control wiper to earthy end at most used
listening setting . This resistor was optional.
SandyK
OP AMP EXPERIENCES AND HOW TO
The initial setup was with a wire link fitted between inverting and output. The optional resistor was suggested for those people who wished to reduce distortion to a minimum. They did however say, that most readers probably wouldn't bother,and would leave the wire link in situ.
SandyK
The initial setup was with a wire link fitted between inverting and output. The optional resistor was suggested for those people who wished to reduce distortion to a minimum. They did however say, that most readers probably wouldn't bother,and would leave the wire link in situ.
SandyK
" ... initial setup was with a wire link fitted between inverting and output. ..." In which case the op-amp is set to unity gain (G = 1) ... and the resistor would "pad" the (-) negative feedback down to a very small fractional gain, which for an input buffer is not good = probably no signal would pass through. (If the volume control pot were, say, 10K, then the resistor should also be 10K = again, unity gain.)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.