Mixi,
Output of this RIAA stage will see variable impedance as the pot is turned. There will be small variation in the characteristics as the pot ist turned. It won't be the problem but it is better if you use P88 after this circuit because in that case RIAA stage will see constant impedance.
Output of this RIAA stage will see variable impedance as the pot is turned. There will be small variation in the characteristics as the pot ist turned. It won't be the problem but it is better if you use P88 after this circuit because in that case RIAA stage will see constant impedance.
quick question regarding the 100nF ceramic caps that are recommended for the power supply. there seem to be two different ways people use them:
1.) connect one cap per opamp across the +/- terminals.
2.) use two per opamp. one from "+" to ground and one from "-" to ground.
any advantages / disadvantages of those approaches?
1.) connect one cap per opamp across the +/- terminals.
2.) use two per opamp. one from "+" to ground and one from "-" to ground.
any advantages / disadvantages of those approaches?
Rod Elliott's Project 88 page says: "Earlier boards included RF interference suppression by adding a small capacitor between the two inputs of U1 (the space for this cap can be seen in the PCB photo, which is of an early board). This has now been removed, as it caused more problems than it solved - in particular, opamp oscillation with some devices." So, I'd be skeptical of #1in this instance.
For #2, I think of small value bypassing capacitors from rail to ground as shunt elements that pass rapidly-varying components of the DC rail voltage to ground. Because the value of the capacitor determines the range of frequencies that will be "bypassed" to ground, sometimes we see bypass networks with a range of capacitor values (10nF, 100nF, 1uF), but usually, a 100nF to ground for each rail is standard and sufficient. Conceptually, #1 is similar, but it prevents rapidly-varying differences between the two rail voltages by passing through high frequency differences from one rail to the other.
-Neil N0FN
For #2, I think of small value bypassing capacitors from rail to ground as shunt elements that pass rapidly-varying components of the DC rail voltage to ground. Because the value of the capacitor determines the range of frequencies that will be "bypassed" to ground, sometimes we see bypass networks with a range of capacitor values (10nF, 100nF, 1uF), but usually, a 100nF to ground for each rail is standard and sufficient. Conceptually, #1 is similar, but it prevents rapidly-varying differences between the two rail voltages by passing through high frequency differences from one rail to the other.
-Neil N0FN
doesn‘t „the two inputs“ refer to the signal in as opposed to the power in?
this is where i got the information for the ceramic caps from: Audio Designs With Opamps
it says: ´but the use of ceramic bypass caps between the supply pins of each device is highly recommended´
.... so that‘s what i did and it seems to work fine. but other people seem to bypass by connecting caps between the supply pins and gnd and i wonder which approach is better?
this is where i got the information for the ceramic caps from: Audio Designs With Opamps
it says: ´but the use of ceramic bypass caps between the supply pins of each device is highly recommended´
.... so that‘s what i did and it seems to work fine. but other people seem to bypass by connecting caps between the supply pins and gnd and i wonder which approach is better?
Hello Mixi,
I always use 0.1uF cap from each supply pin to ground and datasheet recommends it too and generally X7R ceramic are used.
clause 10.2 on page 19 http://www.ti.com/lit/ds/symlink/opa2134.pdf
I always use 0.1uF cap from each supply pin to ground and datasheet recommends it too and generally X7R ceramic are used.
clause 10.2 on page 19 http://www.ti.com/lit/ds/symlink/opa2134.pdf
With LME49720 i use 0,1uf and 10uf cap see datasheet...page 37 http://www.ti.com/lit/ds/symlink/lme49720.pdf
Last edited:
thanks for pointing me in the right direction. i had it wrong on my prototype all along as i used the 0,1uf ceramic to bridge the supply pins. it still worked nicely 🙂. i'll switch to the recommended setup later today and see if it makes any difference.
Mixi,
You did it right. Bypass using single 100nF between supply rails is as good as bypass from supply pins to ground. And is much simpler for DIY-ers, and for PCB design. Doug Self recommends it in his Small Signal Audio book.
You did it right. Bypass using single 100nF between supply rails is as good as bypass from supply pins to ground. And is much simpler for DIY-ers, and for PCB design. Doug Self recommends it in his Small Signal Audio book.
ok, so i guess both options are equally fine. that's good to hear. and i can save 40 cents on buying fewer caps 🙂
Hello Hicoco,
Generally I use whatever available in parts bin, some times I even use film caps in that position.
here is a mouser results, any of these would be ok .
Radial 0.1 uF 100 VDC X7R Multilayer Ceramic Capacitors MLCC - Leaded | Mouser India
regards
Prasi
Do you know which chapter mentions this? I tried to read up on it but couldn‘t find it.
Look at the example Figure 2 - Bypassing The Op Amp Supplies and explanation below it
skrodah, look like the gain on the first stage is 30.6db, i try to figure the second stage... is it between 27,5 and 35db ? if yes how could you have 40db total?The schematic above was updated about four years ago, here's the new one:
.png)
As you know this is the input preamp average or may be that i use the wrong formula.
Thinking quickly: each gain-stage is gain about 30dB, yes.
The in-between R-C network has loss. How much?
It varies with frequency. According to the RIAA curve.
We usually quote phono preamp gain at 1KHz. This type RIAA R-C network has gain of 0 dB at super low frequency, -3dB at 50Hz, -20dB at 1KHz, and -40dB around 20KHz.
So at 1KHz we have gain of (30dB)+(-20dB)+(30dB), which is 40dB, which is very typical.
Switches in the 2nd stage allow you to trim this up or down as needed to suit your system.
hicoco, i believe the poles/zeros are the same relative to each other. bear in mind that pole #1 is usually +20db (you wrote minus). this rc network only attenuates so it can't provide +db. there's +/- 0db at very low frequency and from that point on attenuation increases with frequency.
i guess that's why this kind of network is regarded as not being the best when it comes to noise / headroom. first the signal is amplified, then attenuated, then amplified again.
that being said i use this phono stage quite often and to me it sounds great.
i guess that's why this kind of network is regarded as not being the best when it comes to noise / headroom. first the signal is amplified, then attenuated, then amplified again.
that being said i use this phono stage quite often and to me it sounds great.