Hi,
I will compare the stage when I'm sure I it performs the best it can. That will take some time, I fear.
At the moment, I have a strange phenomenon: I tried the Denon cart DL103-R, and it sounds dull, like underwater. Since it has a lower impedance than the original DL103 (14 vx. 40 Ohms) my guess was, that the original would have lesser trebles, but not so!
I'm totally puzzled!
Rüdiger
I will compare the stage when I'm sure I it performs the best it can. That will take some time, I fear.
At the moment, I have a strange phenomenon: I tried the Denon cart DL103-R, and it sounds dull, like underwater. Since it has a lower impedance than the original DL103 (14 vx. 40 Ohms) my guess was, that the original would have lesser trebles, but not so!
I'm totally puzzled!
Rüdiger
DL-103-R
you are right Rudy! The DL103R is much better than the standard one, it is more open and it is well known that it has a smaller roll-off in the highs compared with the standard.
That's my pickup and I know it very well! You should have experienced a faster bass and tighter bass as well (not night and day... but it is a bit).
The problem is that usually the MC phono stage for such a pickup should be very good and adhoc (well tha's apply to every cart actually).
I use an *old* Burmester 785 (wonderfull German stuffs) which in 1983 implemented the active RIAA... not many gears at time were doing such a big thing.... with an input sensitivity of 0.08mV!... at that time I believe only Counterpoint was able to do that but with passive RIAA!!!
The Burmester is an astonishing preamp the problem is that you never know what cables to use to macth the impedence/capacitance needed. In fact the manual of the preamp has a selection of 10 mc pickups to be used with it... the DENON is not among them. That's way I am following your very interesting post. I am thinking to play around with your circuit.
I have some question:
1. In your very interesting comment you do not discuss the choice of the U3 opamp, why?
2. I am not an expert so I would like to understand (roughly) what is its function;
3. Do you plan to design a bord for this circuit?
4. Which psu are you using to feed it?
By for now
Cukkurullo
Onvinyl said:
you are right Rudy! The DL103R is much better than the standard one, it is more open and it is well known that it has a smaller roll-off in the highs compared with the standard.
That's my pickup and I know it very well! You should have experienced a faster bass and tighter bass as well (not night and day... but it is a bit).
The problem is that usually the MC phono stage for such a pickup should be very good and adhoc (well tha's apply to every cart actually).
I use an *old* Burmester 785 (wonderfull German stuffs) which in 1983 implemented the active RIAA... not many gears at time were doing such a big thing.... with an input sensitivity of 0.08mV!... at that time I believe only Counterpoint was able to do that but with passive RIAA!!!
The Burmester is an astonishing preamp the problem is that you never know what cables to use to macth the impedence/capacitance needed. In fact the manual of the preamp has a selection of 10 mc pickups to be used with it... the DENON is not among them. That's way I am following your very interesting post. I am thinking to play around with your circuit.
I have some question:
1. In your very interesting comment you do not discuss the choice of the U3 opamp, why?
2. I am not an expert so I would like to understand (roughly) what is its function;
3. Do you plan to design a bord for this circuit?
4. Which psu are you using to feed it?
By for now
Cukkurullo
Hi Cukkurullo,
thanks for your interest. If you read the first post, you probably read the posts I linked as well. My own contribution to the circuit appears small then.
U3's purpose is a DC-Servo. It keeps the output at zero Volts DC to avoid the need of a coupling cap. The actual opamp, the AD820, is a textbook choice, Walt Jung and Erno Borbely use it, for example.
I did not try other op's, because I think (well, hope...) its contribution to the sound is small when performing its integrator job well. It is superior to all caps I tried ( I did not try rjm's favorite choice, the black gate).
I will most probably not produce a pcb 'cause I don't plan to make more than one phonopre. If you omit the servo, you might use rjm's board (he is in the process of another batch in the moment, look his thread). I did not try his board, though, and the use of AD811 & AD797 is delicate, so there's a risk.
As a psu I use a super regulator (lots of info about it at diyaudio) which plays a major role for the sound quality.
Yes, I'm aware that the -R performs better. My point ist, that this circuit should perform the better the lower the Z of the cart is. Since the -R *has* this lower Z, the circuit + DL103-R should perform gorgeous together. They just don't.
Rüdiger
thanks for your interest. If you read the first post, you probably read the posts I linked as well. My own contribution to the circuit appears small then.
U3's purpose is a DC-Servo. It keeps the output at zero Volts DC to avoid the need of a coupling cap. The actual opamp, the AD820, is a textbook choice, Walt Jung and Erno Borbely use it, for example.
I did not try other op's, because I think (well, hope...) its contribution to the sound is small when performing its integrator job well. It is superior to all caps I tried ( I did not try rjm's favorite choice, the black gate).
I will most probably not produce a pcb 'cause I don't plan to make more than one phonopre. If you omit the servo, you might use rjm's board (he is in the process of another batch in the moment, look his thread). I did not try his board, though, and the use of AD811 & AD797 is delicate, so there's a risk.
As a psu I use a super regulator (lots of info about it at diyaudio) which plays a major role for the sound quality.
Yes, I'm aware that the -R performs better. My point ist, that this circuit should perform the better the lower the Z of the cart is. Since the -R *has* this lower Z, the circuit + DL103-R should perform gorgeous together. They just don't.
Rüdiger
well ... forget what I've said about 'underwater'. It wasn't the clones' fault. I have new speakers, and the interaction with room and furniture is totally different from my old ones. I should have listened closer while crosschecking with the CD-Player.
Well, the DL103-R plays real good now. I hadn't Problems with to less trebles with the old 103, and I do not notice more treble now. I do notice more resolution now and that hard to describe increase of quality when you get thrilled by that wah-wah guitar, for instance.
Rüdiger
If the feedback resistor R2 stays the same, the gain increases and the bandwidth decreases as you lower the cartridge Z.
Your high frequency cutoff is now three times lower than it was with the 103, assuming you didnt replace R2 with a lower value to match. This is most likely why you perceive less treble. It could also be that the impedance of the cartridge rises more significantly at high frequencies than the 103 did.
-rjm
Hi rjm,
I'm somewhat confused: I've seen you (once?) matched the DL103 with a R2 of 6.8k. I normally use 1.5k both with the DL103 and the DL103-R. I assumed bandwith should be enough, but then, I didn't do the math. With the op27 and the opa627 I tried R2 up to 3k with no audible problems (that with the DL103).
Rüdiger
I'm somewhat confused: I've seen you (once?) matched the DL103 with a R2 of 6.8k. I normally use 1.5k both with the DL103 and the DL103-R. I assumed bandwith should be enough, but then, I didn't do the math. With the op27 and the opa627 I tried R2 up to 3k with no audible problems (that with the DL103).
Rüdiger
It was just a suggestion... at 1.5k, moving from the 103 to the 103R increases the gain of the input stage from 30 to 40 dB. That's still very much within the safe area given a GBWP of 8Mhz (OP27). The 3dB point moves from 240kHz down to 80 kHz, but that does not affect the response at 20kHz significantly. It wouldnt hurt to try 680 ohms though, just to see if there is any difference. If there is none we can put it down to a difference in the impedance characteristics of the two cartridges, which affects the reponse because the gain of the phonoclone has 1/Zcart as a factor.
Hi rjm,
since I'm using the AD811 as frontend, neither a pair of 40/1500 nor of 14/1500 is given in the datasheet. A pair of 51.1/511 (gain of 10) has its -3dB point at 95Mhz. But then... the graph on page 5 quotes ~1.7 Mhz at 1.5k at a gain of +2, who knows how things are at a gain of -107.
So I will check your suggestion.
Rüdiger
since I'm using the AD811 as frontend, neither a pair of 40/1500 nor of 14/1500 is given in the datasheet. A pair of 51.1/511 (gain of 10) has its -3dB point at 95Mhz. But then... the graph on page 5 quotes ~1.7 Mhz at 1.5k at a gain of +2, who knows how things are at a gain of -107.
So I will check your suggestion.
Rüdiger
The AD811 is a high frequency video op-amp. Your bandwidth in any case is up past 40 Mhz or so. All bets are off as to what is actually going on up there, but its safe to say that treble roll-off isnt going to be a problem. 
The AD811 is also a current amplifier, so you can't characterize the bandwidth neatly as a gain-bandwidth product. Although it will never be an issue so far as brute bandwith is concerned, its interesting to note that the output characteristics (including bandwidth but also peaking, overshoot) depend on the absolute value of the feedback resistor (R2) and not just the ratio R2/Zcart. It looks like quite a large difference occures over the range 500 to 1500 ohms (Fig 11, 12). I would have thought that large=better in your case, since only audio frequencies are of interest, but you never know.
-rjm
The AD811 is also a current amplifier, so you can't characterize the bandwidth neatly as a gain-bandwidth product. Although it will never be an issue so far as brute bandwith is concerned, its interesting to note that the output characteristics (including bandwidth but also peaking, overshoot) depend on the absolute value of the feedback resistor (R2) and not just the ratio R2/Zcart. It looks like quite a large difference occures over the range 500 to 1500 ohms (Fig 11, 12). I would have thought that large=better in your case, since only audio frequencies are of interest, but you never know.
-rjm
Hi rjm,
late last night I tried a R2 of ~860R, and the sound changed drastically to the bright side! At the same time, it seems quiet independent of injected noise, you hardly hear it when you touch the input with disconnected cart! I tend to agree, that the absolut value of R2 plays a major role, something I did not take into account since now.
It seems to be worth the effort to finetune R2. I seriously doubt now, that you can easily switch cartridges without changing R2 as well.
It would be interesting to see a impedance vs. frequency plot of the 103-R (as you have it for speakers).
Interestingly, the frequency measurement provided with the 103-R is made with an input-R of 47k (instead of 1k with the standard-103), that may be a hint that your mentioned impedance dependency is indeed true.
Rüdiger
late last night I tried a R2 of ~860R, and the sound changed drastically to the bright side! At the same time, it seems quiet independent of injected noise, you hardly hear it when you touch the input with disconnected cart! I tend to agree, that the absolut value of R2 plays a major role, something I did not take into account since now.
It seems to be worth the effort to finetune R2. I seriously doubt now, that you can easily switch cartridges without changing R2 as well.
It would be interesting to see a impedance vs. frequency plot of the 103-R (as you have it for speakers).
Interestingly, the frequency measurement provided with the 103-R is made with an input-R of 47k (instead of 1k with the standard-103), that may be a hint that your mentioned impedance dependency is indeed true.
Rüdiger
Decompensation cap values
Dear Onvinyl,
You did not specifiy the value you used for the decompensating caps you used with the AD811. What did you use?
I am determined to try the LME49710's but with rjm's concern that it may be too fast I am thinking that decompensation might be the key for the first stage.
Would the scheme you used work with the LM#49710 (Op amps are still mysteries to me, please forgive me)?
Your help is appreciated,
Dear Onvinyl,
You did not specifiy the value you used for the decompensating caps you used with the AD811. What did you use?
I am determined to try the LME49710's but with rjm's concern that it may be too fast I am thinking that decompensation might be the key for the first stage.
Would the scheme you used work with the LM#49710 (Op amps are still mysteries to me, please forgive me)?
Your help is appreciated,
Re: Decompensation cap values
Are you referring to the decoupling caps?
I use Panasonic FC 100µ here, and paralleled with some foils caps around 100nF. Do you have a scope? You need to check whether parallel caps improve or worsen the op-amps behaviour and if your decoupling works as expected.
The key lies in a most optimized physical layout of the circuit. This is in my view the heart of the art of building circuits, and, sadly, I am by no means an expert in this area. I'm just trying things over and over to achieve some decent performance. There is a lot of info and stuff to find on this topic here at the forum. (Tight layout, short signal path, proper grounding, current returns and the like.).
Playing and testing various decoupling schemes comes after the layout, you almost always can't save a 'faulty' layout just by popping in different caps.
After all, I did not succeed in having both the AD797 (radio) and the LT1028 (fancy noises) in the first stage, this is probably because the cartridge is in a way a part of the feedback loop -> byebye short signal path.
The AD811 is suprisingly tame in this position, ask rjm if he tried this with his boards if you use them.
I would guess that your chip might work well in the second stage. I would have tried, but I did not receive my samples.
hope that helps,
Rüdiger
rickmcinnis said:
Are you referring to the decoupling caps?
I use Panasonic FC 100µ here, and paralleled with some foils caps around 100nF. Do you have a scope? You need to check whether parallel caps improve or worsen the op-amps behaviour and if your decoupling works as expected.
rickmcinnis said:
The key lies in a most optimized physical layout of the circuit. This is in my view the heart of the art of building circuits, and, sadly, I am by no means an expert in this area. I'm just trying things over and over to achieve some decent performance. There is a lot of info and stuff to find on this topic here at the forum. (Tight layout, short signal path, proper grounding, current returns and the like.).
Playing and testing various decoupling schemes comes after the layout, you almost always can't save a 'faulty' layout just by popping in different caps.
After all, I did not succeed in having both the AD797 (radio) and the LT1028 (fancy noises) in the first stage, this is probably because the cartridge is in a way a part of the feedback loop -> byebye short signal path.
The AD811 is suprisingly tame in this position, ask rjm if he tried this with his boards if you use them.
I would guess that your chip might work well in the second stage. I would have tried, but I did not receive my samples.
hope that helps,
Rüdiger
Another faux pas!
Onvinyl,
I looked at the application note and SHOULD have figured out that these were decoupling caps. That explains why I found it all so confusing. Very embarrassing!
Do you think slowing down the op amp will help? I am planning on having this box within 12 inches of the cartridge. I know that is still a long path, but I wonder if it would help?
I have four of the LME49710's on the way (in TO-99 package).
If you would like to try them I would be glad to donate two to the cause. DIGIKEY is stocking them. It is easy enought to get more if something good happens.
Send me an email of there is any interest.
Thanks for your help and advice,
Onvinyl,
I looked at the application note and SHOULD have figured out that these were decoupling caps. That explains why I found it all so confusing. Very embarrassing!
Do you think slowing down the op amp will help? I am planning on having this box within 12 inches of the cartridge. I know that is still a long path, but I wonder if it would help?
I have four of the LME49710's on the way (in TO-99 package).
If you would like to try them I would be glad to donate two to the cause. DIGIKEY is stocking them. It is easy enought to get more if something good happens.
Send me an email of there is any interest.
Thanks for your help and advice,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.