questions
First let me say I am not a whiz in the electronics. Especially in the digital world. I have the double stacked AD844 I/V system in place and I am using the in chip buffer. I have used the single and double output buffer connection and there is a difference. Presently I have the null circuit that George uses and have nulled out the buffer output to basically zero on the output. I also have a 2.7k/470pf filter on the TZ pin 5. I installed this back in the threads first couple hundred posts and now have gone back an re-read the entire thread. Now I have a few questions.
1. If we are using the 844 in the inverting mode, is the null feature of this chip non usable?
2. Is it important to null out pin 5 if we are using the in chip buffer as there is an offset if the output buffer is nulled?
3. If I install the large amount of bypass caps between +/- rails, would the two ceramic and silmic caps on the +/- rails to ground pass the high freq noise to ground planes and upset the output of the buffer or?
4. My amps have a built in volume control, a 250k pot. I have plenty of gain with the 2.7k TZ resistor. My question is if I lowered this value to the minimum(say 1.5k) and with the two TZ pin in parallel, would the output impedance be low enough to put directly into the 250k input impedance of the amps with little or no adverse effects? I saw George's comment on loading down this pin 5 output with 100k, but how about 1.5k in parallel with two pin 5s into a 250k.
Also, since the resistor on pin 5 is close to the source of the signal, it does play a part in the sound. I replaced an ordinary MF with a Mepco RN MF and found a change for the better. Don't want to start a new topic, but it also has an impact on the sound.
First let me say I am not a whiz in the electronics. Especially in the digital world. I have the double stacked AD844 I/V system in place and I am using the in chip buffer. I have used the single and double output buffer connection and there is a difference. Presently I have the null circuit that George uses and have nulled out the buffer output to basically zero on the output. I also have a 2.7k/470pf filter on the TZ pin 5. I installed this back in the threads first couple hundred posts and now have gone back an re-read the entire thread. Now I have a few questions.
1. If we are using the 844 in the inverting mode, is the null feature of this chip non usable?
2. Is it important to null out pin 5 if we are using the in chip buffer as there is an offset if the output buffer is nulled?
3. If I install the large amount of bypass caps between +/- rails, would the two ceramic and silmic caps on the +/- rails to ground pass the high freq noise to ground planes and upset the output of the buffer or?
4. My amps have a built in volume control, a 250k pot. I have plenty of gain with the 2.7k TZ resistor. My question is if I lowered this value to the minimum(say 1.5k) and with the two TZ pin in parallel, would the output impedance be low enough to put directly into the 250k input impedance of the amps with little or no adverse effects? I saw George's comment on loading down this pin 5 output with 100k, but how about 1.5k in parallel with two pin 5s into a 250k.
Also, since the resistor on pin 5 is close to the source of the signal, it does play a part in the sound. I replaced an ordinary MF with a Mepco RN MF and found a change for the better. Don't want to start a new topic, but it also has an impact on the sound.
Hi.....
...just have see this trade and becouse Im in to building my dac with tda1541a dac chip I will have some questions.Plese tell me is I go with Pedja discrete ad844 opamp for I/V output stage and if I decide to stack up three opamps per chanel must I change the resistor on pin 5 and the capacitor or I can stay with R=1,5kOhm an C=1nF ?And another question,plese tell me if I decide to go with external buffer BUF03 from AD and I take the signal from pin 5 on AD844 opamp,must I than change any parts on the circuit or no??
...just have see this trade and becouse Im in to building my dac with tda1541a dac chip I will have some questions.Plese tell me is I go with Pedja discrete ad844 opamp for I/V output stage and if I decide to stack up three opamps per chanel must I change the resistor on pin 5 and the capacitor or I can stay with R=1,5kOhm an C=1nF ?And another question,plese tell me if I decide to go with external buffer BUF03 from AD and I take the signal from pin 5 on AD844 opamp,must I than change any parts on the circuit or no??
As far as I read on this thread 1K5 resistor and 1nF cap remain the same regardless the number of AD844s. Should you need a higher gain you can use 1K8 or 2K2.
You can get signal on PIN 5 or at IC output pin. You can try, but it's safer to use a buffer. One only might be better, so don't connect those pins together.
You can get signal on PIN 5 or at IC output pin. You can try, but it's safer to use a buffer. One only might be better, so don't connect those pins together.
Hi.....
...just have see this trade and becouse Im in to building my dac with tda1541a dac chip I will have some questions.Plese tell me is I go with Pedja discrete ad844 opamp for I/V output stage and if I decide to stack up three opamps per chanel must I change the resistor on pin 5 and the capacitor or I can stay with R=1,5kOhm an C=1nF ?And another question,plese tell me if I decide to go with external buffer BUF03 from AD and I take the signal from pin 5 on AD844 opamp,must I than change any parts on the circuit or no??
That should be possible if buffer's input impedance is high. 200K or more, but that shouldn't be a problem. Anyway I'd listen both pin 5 and buffered output.
If tube preamp also has gain I'd be possible to install a low value resistor. I will soon test this solution. DC couple AD844 and tube preamp. Should you use a coupling capacitor you might get better results by internal IC buffer....
If tube preamp also has gain I'd be possible to install a low value resistor. I will soon test this solution. DC couple AD844 and tube preamp. Should you use a coupling capacitor you might get better results by internal IC buffer....
Hi Fralippo.......
....thanks for the answer.Now my question is if can use tube buffer on the output of the pin 5?I have some realy nice buffer with e88cc tubes hat I dont use anymore.So what about that????
Last edited:
.....thanks again for the answer.Yes my tube buffer is constructed with e88cc tubes and have the gain of 1,so practicaly speaking what is coming in is going out,so there is no attenuation only the tube sound is aded to the output signal.The input impedance of my tube buffer is 470kohm,so there shuld be no problem,like you say.But in any case tellme if I must change some parts around AD844 or no.I just think to use output on pin 5 and than put this on the input of the buffer where I have 0,22uf input capacitor Mundor Silver oil in this case.
What do you say about this solution???
What do you say about this solution???
You'd check if there's DC between buffer's tube input grid and ground. It'd probably be, beeing a buffer. Try to post schematic. I much love tubes too..... but If you've to use two caps.... even good ones, I'd surely use AD844 buffer's DC coupled to amp! You can bet it'd sound better.
Concentrate instead on seriously supply +-15V to IC. Differences are heavy here. I always had very good results with ALWSR regulators (Walt Jung).
Concentrate instead on seriously supply +-15V to IC. Differences are heavy here. I always had very good results with ALWSR regulators (Walt Jung).
...Hi,yes I understand what you mean.The schematic is simply build and cloned from the musical fidelity x10d buffer.I made this some year ago but with verry,verry good parts on it like shinkoh resistors,mundorf silver oil coupling caps and jupiter beeswax output caps,but now I dont use this anymore.For power supply I just have at home from m past project a pair of -12 and +12 volts regulators fro Dexa.This are ultra low noise regulators.So I will just build some good psu with toroid transformer with mundorf Ag caps(I have this at home)than the two Dexa UWB regulators to power the AD844 and BS245A.Im just courius what will be out of this project,becouse some old Philips cd that have tda1541a chip from my friend have get my atention,the player have some mod inside,and it sounds so good against my audioaero prima cd player,that I now buy some rotel rcd855 cd player with the same dac inside I just want to see what is posibile to do with some simple modification to become bether sound out of the rotel unit against my audioaero cd player.Its just this simple natural sound that shocked me from this old scool tda philips chip.
those old dac chips are multibit.... that's why they're so sonically superior!
Feed the 1541 with Dexa or similar regulators too. Big improvements. +-5V and -15V.
Replace clock also. And check if you'd fit a PDM100 digital filter in place of the original one (SAA7220 ?).
+-15V on AD844s might be better than +-12.
I'd avoid toroid transformers for audio.
Feed the 1541 with Dexa or similar regulators too. Big improvements. +-5V and -15V.
Replace clock also. And check if you'd fit a PDM100 digital filter in place of the original one (SAA7220 ?).
+-15V on AD844s might be better than +-12.
I'd avoid toroid transformers for audio.
.....I just will use this parts becouse I have this at home and just to try out what is posible to build and improved the sound of tda1541.Plese tell me how to do the mod with pmd100 insteed of sa7220.I never think about this.I have already instal the tentlabs xo3 clock and put bether decoupling caps for the tda1541 dac chip and I replace all the psu caps for better.Like I say,if the results will be good,than I will do some mods to the other part of the player allso.Im not such good experts in electronics,I have build some 50 pairs of verry good loudspeakers,and build a few verry simple tube preamps and some gainlcones,but thats it.So must learn a lot of new stuf in moding the cd player.
After doing Abraxalito's mod with the 10uf smd cap from + to - on the 2 stack AD844 which is a very nice improvement.
I thought I wonder what the internal buffer of the 844 will sound like now, after all the cap mod should effect this as well, compared to what I've been using, the BUF03 which I'm very happy with.
I've just finished going back to the 2 x stack AD844's internal buffer I will report back with the sound, after I've listened for a while.
If you decide to use one of the internal buffers in a stacked 844 I/V arrangement, you can then use the dc null offset pins (1&8) only on the 844 that your using the buffer of.
And there is a misprint on the AD844 data sheet the dc offset trimpot should be 20kohm not 20ohms, Charles Hansen also found this out with the Ayre CX-7ex dac and CX-7e cdp where he uses 8 x AD844's.
Back to you later with the sound of the internal buffer.
Cheers George
I thought I wonder what the internal buffer of the 844 will sound like now, after all the cap mod should effect this as well, compared to what I've been using, the BUF03 which I'm very happy with.
I've just finished going back to the 2 x stack AD844's internal buffer I will report back with the sound, after I've listened for a while.
If you decide to use one of the internal buffers in a stacked 844 I/V arrangement, you can then use the dc null offset pins (1&8) only on the 844 that your using the buffer of.
And there is a misprint on the AD844 data sheet the dc offset trimpot should be 20kohm not 20ohms, Charles Hansen also found this out with the Ayre CX-7ex dac and CX-7e cdp where he uses 8 x AD844's.
Back to you later with the sound of the internal buffer.
Cheers George
Initial impressions are it's a very good buffer, I would say better that most "hifi" opamp buffers that use feedback, even the expensive ones.
So if you have no buffers better than these, go for the zero feedback internal buffer of the AD844.
PS: you can trim the dc offset of everything including the dac's offset down to +-1mv with a 20kohm pot on pins 1 & 8.
As for how it sounds agaist the BUF03 same gain, I'll listen some more but so far I would say the BUF03 has an "awe" "bigness" and power factor that the AD844's buffer seems to lack a little of so far.
BTW: Does any manufacturer make a buffer these days with fet input, bi-polar output, and with no feedback like the BUF03???
Cheers George
So if you have no buffers better than these, go for the zero feedback internal buffer of the AD844.
PS: you can trim the dc offset of everything including the dac's offset down to +-1mv with a 20kohm pot on pins 1 & 8.
As for how it sounds agaist the BUF03 same gain, I'll listen some more but so far I would say the BUF03 has an "awe" "bigness" and power factor that the AD844's buffer seems to lack a little of so far.
BTW: Does any manufacturer make a buffer these days with fet input, bi-polar output, and with no feedback like the BUF03???
Cheers George
Last edited:
I don't know of an integrated buffer but I'm building a discrete buffer just like you describe it. FET input, bipolar output, no (global) feedback. Its using a JFET, an emitter follower and two current sources, so not too hard to build. However its not DC accurate. Will report back when I've listened.....
Suggest you try increasing the decoupling on the AD844, and try loading it with a CCS to get more 'awe' factor.
Suggest you try increasing the decoupling on the AD844, and try loading it with a CCS to get more 'awe' factor.
Will you be posting this up abraxilito if all works well, does it need a coupling cap or can it be trimed out for small amount of offset when settled, as my set up mutes the output on and off for a few seconds while everything stablizes.
Cheers George
No buffer connection
I have connected pin 5 directly to the amp input (250k pot), see above question on my last post, bypassing the single connected internal buffer on the stacked AD844s. I have only listened for a half hour, but there is no going back. An overall improvement, especially in detail, dynamics and smoothness. Like George said earlier, you just want to continue to raise the volume without limit. Much of the edge of the digital sound disappears and in my system the mids came out to my satisfaction which I have been trying to increase for some time. I don't sense a loss in low freq response (do to impedance mismatch) but have not made any measurements. I like to keep things as simple as possible and removing the buffer from the system fits my philosophy. Player, amp and speaker, all directly connected (soldered connections) with only one cap in the signal path. NICE.
I have connected pin 5 directly to the amp input (250k pot), see above question on my last post, bypassing the single connected internal buffer on the stacked AD844s. I have only listened for a half hour, but there is no going back. An overall improvement, especially in detail, dynamics and smoothness. Like George said earlier, you just want to continue to raise the volume without limit. Much of the edge of the digital sound disappears and in my system the mids came out to my satisfaction which I have been trying to increase for some time. I don't sense a loss in low freq response (do to impedance mismatch) but have not made any measurements. I like to keep things as simple as possible and removing the buffer from the system fits my philosophy. Player, amp and speaker, all directly connected (soldered connections) with only one cap in the signal path. NICE.
The output from Pin 5 is actually current (very high Z) and the output of a transconductance amplifier. Whatever value you use on Pin 5 and the 250K pot are in parallel, so the gain will go down fractionally. Effectively the final R value is a passive I/V converter and now the final output Z is determined by those resistors rather than the high Z of Pin 5. You can measure the output Z by measuring the signal before and after the additional 'load' of 250K, and if Pin 5 value R is low, then the output Z may be reasonably low at that. Many tube output stages that many love are not that low Z and you may well be in the same ballpark and getting away with it.
Also, the the trimpot on Pins 1/8 that is intended to null the the DC output of the buffer will also work on nulling Pin 5. Then ignore the DC on the buffer (Pin 6) which may have some 10-20mV (?) difference. This way no cap on Pin 5 may be needed. But make sure that the 250K load is seen by Pin 5 when doing the DC nulling/trimming and don't change it (keep hardwired will help). Keep capacitance low or the extreme bandwidth available on Pin 5 will be rolled off significantly.
No, I am not surprised that it sounds good.
Cheers, Joe
.
This will work fine if there is no resistor to ground after the 250k pot on your amp before the input tube/transistor, if there is then you will most probably restrict the dynamics, as the TZ will then see the combination of the 250k pot + the resistor. And as I found if it (the TZ)sees anything 100kohm or less will cause a bit of a bottleneck for the signal, and start to attenuate it, which to my ear caused a dynamic restriction, even if the volume was raised to the same level.
Cheers George
Last edited:
I have now listened all weekend to the internal buffer, and let me tell you it is very good. Use it if you want to keep it simple, as I think it is far better than any external hifi opamp feedback buffer I've heard.
But it still does not better the BUF03, maybe it's the massive class A it's biased into or it's current with 1ohm output impedance, that gives it a meaty-ness to the sound that ust sounds so right, and it has a "shock awe" factor that the internal buffer can't match, and a naturalness that is just right also so right.
You can follow any instrument or vocal even though it's surrounded by much louder things, like there is a greater sense of space between everything.
Cheers George
But it still does not better the BUF03, maybe it's the massive class A it's biased into or it's current with 1ohm output impedance, that gives it a meaty-ness to the sound that ust sounds so right, and it has a "shock awe" factor that the internal buffer can't match, and a naturalness that is just right also so right.
You can follow any instrument or vocal even though it's surrounded by much louder things, like there is a greater sense of space between everything.
Cheers George