Grid to ground i/v works well Walter.. OMG... your right again.
Are your sure your not Fikus pretending to be Walter ?
Thanks for this additional very noticeable step - I'm running op amp free at last !!
Now I seem to have even more detail, more sharper edges, extra slam and ever more bass definition.
It's kind of embarrassing to read that lot back as I've always been a little bit of a sceptic reading other peoples ' review style ' raptures.
I may well become a valve convert.
Power supply only a few days away and then I may just stop altogether.
I can hardly believe how nice it's become all of a sudden.
When you look at ecdesigns and Thorstens approach to 1541 application it makes you wonder what they are hearing compared to our humble attempts. Their machines must be mind bogglingly fantastic.
I'll be happy to stop soon - chasing the ultimate is frankly beyond my level of understanding and technical capability - and I'm not ashamed to say that.
Some super regs and maybe, and I mean maybe...i2s mods and that's it.
Even last week, before all this i/v stuff my son came down the stairs to ask me to stop playing the piano - he was on the phone and couldn't properly hear his mate speaking......I wasn't playing the piano at all !!
What sort of compliment to my system is that ?
I'll take it
Andrew
Are your sure your not Fikus pretending to be Walter ?
Thanks for this additional very noticeable step - I'm running op amp free at last !!
Now I seem to have even more detail, more sharper edges, extra slam and ever more bass definition.
It's kind of embarrassing to read that lot back as I've always been a little bit of a sceptic reading other peoples ' review style ' raptures.
I may well become a valve convert.
Power supply only a few days away and then I may just stop altogether.
I can hardly believe how nice it's become all of a sudden.
When you look at ecdesigns and Thorstens approach to 1541 application it makes you wonder what they are hearing compared to our humble attempts. Their machines must be mind bogglingly fantastic.
I'll be happy to stop soon - chasing the ultimate is frankly beyond my level of understanding and technical capability - and I'm not ashamed to say that.
Some super regs and maybe, and I mean maybe...i2s mods and that's it.
Even last week, before all this i/v stuff my son came down the stairs to ask me to stop playing the piano - he was on the phone and couldn't properly hear his mate speaking......I wasn't playing the piano at all !!
What sort of compliment to my system is that ?
I'll take it
Andrew
Excellent!! I am pleased it worked well.
That is what I love about this community of interest. Fikus got me interested in the TDA1541a on tubes, you got me to do the DEM upgrade, Simon got me onto super regs, Peter D took care of amp and I2S, CMP radical digtal transport and guys like EC design & NelsonP spur us all on... It is a great support system. Oh and how could I forget Bud Purvine for showing me how to transform 30 year old speakers into better than I could afford to buy?
Your son's comment says it all. I have the same reaction when I am a couple of rooms away. It used to sound like a stereo playing. Now it sounds like some jazz band is playing down the hall.
I must ask... It sounds like you are successful using the 6n6p tubes. Is that correct? If so, that is an important step. Now you have me thinking of upgrading the filament power supply to allow switching them in... I find that by toggling between two resistor values in the CRC filter for filament, you can roll tubes between 6np6 on one setting and the 6n2p/6n1p crowd on the other. I think the 6n6p is a little more desireable than the 6n2p based on my experience on the Sony. Don't feel you need to switch away from these. IMO, they are the top of the hill. They have a smooth natural sound that is delightful. The 6n2p is slightly more spacy in sound stage, but somehow a little contrived by comparison.
I too am considering this good enough. I will continue to tinker with power supply on my USB to IS2 board. My main project now returns to the amps so that dpt can catch up to the TDA1541a. One day speakers.. such fun.
decoupling with surface mount organic polymers
Now this one falls in the land of finishing touches, but its worth noting. Again from the prolific research by EC Design, I was intrigued by his approach to decoupling the tda1541a with surface mount caps. He claims every mm counts and he puts caps directly on pins 7-13 and 18-24.
I picked up some surface mount 8u organic polymers to experiment here. I put one right on the power pin of my clock. Very nice jump in bass, texture etc. (still no super regs so you may have this, but the cap alone worked for me).
So I thought what the heck, lets try the decoupling. I already have BGN on the all important pins 13 and 18, so should already be good, right?. Now under the PCB there is no room with all my DEM stuff. So why not right on the pins as they come from the chip top side. First thing you see is the tip of the soldering iron is way bigger than the cap. And it is already tight with all these caps mounted by Arcam. Never say die.. I put some 20 guage copper wire wrapped around a 60w iron and left a piece sticking out to act as a tiny iron. Put a good sized tin of solder on both the cap and the pin and then with tweezers, a lot of swearing and failed attemps finally got these things to stick on the pins. Then separate wires to ground.
Hooked it up and I must say it is good. At first a little bassy and congested. Of course new wire and new caps always do that. Now with an hour on it, it's opening up really nicely. Nothing of a holly **** you gotta do it, but a nice relaxed polish. Kind of like the clear coat on a Jaguar. Its already got lots of great paint, and its a beauty. But the clear coat just finishes it. Lots of texture. Word that springs to mind is coherent.
It is seriously a pain in the behind to solder. If I ever build a tda1541a dac from scratch, I'm using these caps on the pins. For now, likely just on a couple.
cheers
Now this one falls in the land of finishing touches, but its worth noting. Again from the prolific research by EC Design, I was intrigued by his approach to decoupling the tda1541a with surface mount caps. He claims every mm counts and he puts caps directly on pins 7-13 and 18-24.
I picked up some surface mount 8u organic polymers to experiment here. I put one right on the power pin of my clock. Very nice jump in bass, texture etc. (still no super regs so you may have this, but the cap alone worked for me).
So I thought what the heck, lets try the decoupling. I already have BGN on the all important pins 13 and 18, so should already be good, right?. Now under the PCB there is no room with all my DEM stuff. So why not right on the pins as they come from the chip top side. First thing you see is the tip of the soldering iron is way bigger than the cap. And it is already tight with all these caps mounted by Arcam. Never say die.. I put some 20 guage copper wire wrapped around a 60w iron and left a piece sticking out to act as a tiny iron. Put a good sized tin of solder on both the cap and the pin and then with tweezers, a lot of swearing and failed attemps finally got these things to stick on the pins. Then separate wires to ground.
Hooked it up and I must say it is good. At first a little bassy and congested. Of course new wire and new caps always do that. Now with an hour on it, it's opening up really nicely. Nothing of a holly **** you gotta do it, but a nice relaxed polish. Kind of like the clear coat on a Jaguar. Its already got lots of great paint, and its a beauty. But the clear coat just finishes it. Lots of texture. Word that springs to mind is coherent.
It is seriously a pain in the behind to solder. If I ever build a tda1541a dac from scratch, I'm using these caps on the pins. For now, likely just on a couple.
cheers
Attachments
I may be sick but...
I had these 14 organic polymer surface mount caps now sitting in the drawer, so rather than sit and enjoy listening to music I was compelled to put them in.
Attached pic for the record shows 6 on each side on the bottom. The first one on each row is above with an even shorter path from pin to cap.
The ground is 24 guage copper in teflon. Each side separately grounded direct to star ground on tube output, so there is no interferance with any other ground.
I hate to say it, but if you ever get in to this area of your machine, its worth considering. It will likely get more open and airy after a few hours on new caps. It is very quiet and easy right now. Well worth the squinting with tweezers to place these little devils without messing up. You can see my poor BGN and Mcap took a beating from getting touched by the hot iron.
I had these 14 organic polymer surface mount caps now sitting in the drawer, so rather than sit and enjoy listening to music I was compelled to put them in.
Attached pic for the record shows 6 on each side on the bottom. The first one on each row is above with an even shorter path from pin to cap.
The ground is 24 guage copper in teflon. Each side separately grounded direct to star ground on tube output, so there is no interferance with any other ground.
I hate to say it, but if you ever get in to this area of your machine, its worth considering. It will likely get more open and airy after a few hours on new caps. It is very quiet and easy right now. Well worth the squinting with tweezers to place these little devils without messing up. You can see my poor BGN and Mcap took a beating from getting touched by the hot iron.
Attachments
Polymer Aluminum Electrolytic Capacitor
I got them at Future Electronics. $0.59 each. They are:
CORNELL DUBILIER
ESRD Series 7.3 x 4.3 mm 16 V 8.2 uF Polymer Aluminum Electrolytic Capacitor
Mfr Part#: ESRD8R2M16R
Mounting Method: Surface Mount
Packaging: Cut Tape
Walter
Type: Solid Chip
Rated Voltage: 16 V
Capacitance: 8.2 µF
Case Size/Dimension: 7.3 x 4.3 mm
Jon,
I got them at Future Electronics. $0.59 each. They are:
CORNELL DUBILIER
ESRD Series 7.3 x 4.3 mm 16 V 8.2 uF Polymer Aluminum Electrolytic Capacitor
Mfr Part#: ESRD8R2M16R
Mounting Method: Surface Mount
Packaging: Cut Tape
Walter
Type: Solid Chip
Rated Voltage: 16 V
Capacitance: 8.2 µF
Case Size/Dimension: 7.3 x 4.3 mm
Thanks Walter, your results are interesting. I'm not sure if anyone has tried such high values on all pins before. I was planning on trying some PPS in the 10-100s nF range when I get the chance. I did try Oscons on the MSB but the leakage is rather high so I went back to film. SMT would be better though.
re leakage
Jon
I'll need to let these fully settle before I comment on the sound. My impression after a couple of hours on the first 2 pins was favorable enough to motivate me to do the rest. I am no expert at critical listening.. One initial impression that ran through my mind after a couple of hours was prior to this change all sounds were somehow linked.. like in a 3d spider web. The impression is the web is now gone and sounds stand on their own with their own texture and harmonic. Not isolated but somehow distinct. I know that sounds wierd but that was the impression. Certainly no 'digital' edge.
What is the audible impact of leakage? The original 220n film are still there technically acting as bypass to the 8.2u organics.
Jon
I'll need to let these fully settle before I comment on the sound. My impression after a couple of hours on the first 2 pins was favorable enough to motivate me to do the rest. I am no expert at critical listening.. One initial impression that ran through my mind after a couple of hours was prior to this change all sounds were somehow linked.. like in a 3d spider web. The impression is the web is now gone and sounds stand on their own with their own texture and harmonic. Not isolated but somehow distinct. I know that sounds wierd but that was the impression. Certainly no 'digital' edge.
What is the audible impact of leakage? The original 220n film are still there technically acting as bypass to the 8.2u organics.
Is this what your referring too Walter ?
I was a bit baffled by this picture attached but see the sense in the way you've done yours with regard grounding to the star elsewhere - might even have a go at this.
If pin 1 is the bottom right on this attachment why is pin 4 ( Data R ) and 5 ( Analogue Ground ) shorted ?
I've got a clock feed going to pin 4 so this has completely thrown me
Not questioning ecdesigns - just my own grasp of it all
Thanks
Andrew
I was a bit baffled by this picture attached but see the sense in the way you've done yours with regard grounding to the star elsewhere - might even have a go at this.
If pin 1 is the bottom right on this attachment why is pin 4 ( Data R ) and 5 ( Analogue Ground ) shorted ?
I've got a clock feed going to pin 4 so this has completely thrown me
Not questioning ecdesigns - just my own grasp of it all
Thanks
Andrew
Attachments
Hi Walter, sorry I hadn't appreciated that the original caps were still there, that makes a difference !
Regarding the leakage 4mA is the DACs full scale current so divide that by 65536 and you get 61nA so a cap with a leakage current of up to 3uA is significant. You'd should expect to hear the difference in quiet passages where the signal is only using the lower bits.
Jon
Regarding the leakage 4mA is the DACs full scale current so divide that by 65536 and you get 61nA so a cap with a leakage current of up to 3uA is significant. You'd should expect to hear the difference in quiet passages where the signal is only using the lower bits.
Jon
The key to the use of pin 4 is actually pin 27. It sets the mode the 1541 expects on input. pin 27 is set to +5v by jumping 27 to 28. This sets the input mode to "time MUX TWC". In this case L&R channel signals come in multiplexed on pin 3. Pin 4 is not used. Looking closely at the pic of my board, you will see pin 4 is shorted to ground. I have reinforced it by running a copper wire from pin 4 over to output ground. I connect to analogue ground en route with the big BGN cap.
Jon, thank you for this explanation. It is an area I need more reading to fully appreciate.
I confess I did not take great care in the cap selection. ECDesigns uses 1u film caps. I was quite impressed by the OSCON SEPs so I took a leap of faith by choosing the 8.2u SMT organics.
They likely are not the optimum choice and I welcome the experimental findings of others in this regard. I think they are a step in the right direction though. Early on, I replaced the caps at 13 & 18 with small BGN as someone had reported good result. To me it made no real gain. When I added these SMT organics right on the pin, however the result was significant. Adding the rest is a marginal improvement but an improvement at that. In this application they are the much sought after BG killer.
Another application with similar result was my clock PS. My clock is a surface mount device originally configured with a 10u oscon right on the power pin. I added one of these 8.2u organics soldered directly to the power pin. Substantial improvement in sound quality.
Last edited:
Early listening impressions very good
First, an appology to Andrew's wife. I know I am going to interest you in this and for that I send my appologies to your sweetie. You are soon going to be hold up in the basement cursing me for making you solder these little things in such a tight space.
I let it run for a couple of hours before listening. All told there are 4-5 hours on the new caps and the sound has exploded. I don't know how universal this is, but on my machine it is substantial. Lukas talks about eargasms, goosebumps and getting startled in the dark listening to a great machine. I am getting all that. These caps have somehow tied together all we have been doing and brought it all together. I wish I had the audio vocabulary to describe it. Each instrument stands alone and it resonates. Yet they still tie together coherently. There is a texture on strings that you feel as much as hear. The bells, triangles etc just are natural. Pianos, drums and cymbals are all just real. Quite alot of micro information is definitely having an effect that makes you feel the music more than conciously listen to it. My toe taps. I wanted to get up and dance. Trust me I am not a dancer, so I know there is something working at a low level.
Under $10 and about 2 hours. This is worth it.
First, an appology to Andrew's wife. I know I am going to interest you in this and for that I send my appologies to your sweetie. You are soon going to be hold up in the basement cursing me for making you solder these little things in such a tight space.
I let it run for a couple of hours before listening. All told there are 4-5 hours on the new caps and the sound has exploded. I don't know how universal this is, but on my machine it is substantial. Lukas talks about eargasms, goosebumps and getting startled in the dark listening to a great machine. I am getting all that. These caps have somehow tied together all we have been doing and brought it all together. I wish I had the audio vocabulary to describe it. Each instrument stands alone and it resonates. Yet they still tie together coherently. There is a texture on strings that you feel as much as hear. The bells, triangles etc just are natural. Pianos, drums and cymbals are all just real. Quite alot of micro information is definitely having an effect that makes you feel the music more than conciously listen to it. My toe taps. I wanted to get up and dance. Trust me I am not a dancer, so I know there is something working at a low level.
Under $10 and about 2 hours. This is worth it.
I stand corrected..
I took another look at the ARCAM service manual. They have 27 set to +5v meaning 4 is not used. They have a jumper from 2 to 4 so they are both tied to clock.
EcDesigns has wired 4 to ground. Since it is not used, I guess it's 2 different ways of dealing with it by 2 different engineers.
I took another look at the ARCAM service manual. They have 27 set to +5v meaning 4 is not used. They have a jumper from 2 to 4 so they are both tied to clock.
EcDesigns has wired 4 to ground. Since it is not used, I guess it's 2 different ways of dealing with it by 2 different engineers.
Walter Hi
Wifey said the apology was NOT accepted
( she laughed )Seems you've gone further than anyone with cap values and found an improvement - irrespective of the leakage current issues - bravo !
Are we back to test meter results versus the ears again maybe ?
I'll stay well out of that debate but would like to try your version.
I'm considering trying NOS too but as yet am unsure of how to do it. I did it on the Alpha 5 but it was easy as the 7220 is right next to the 1541A.
I lifted data, bck and ws resistors and attached wires to either end of them. Then simply connected the ends of these now extended resistors to pins 1,2 and 3 of each chip ( in the correct order of course ! )
Worked first time.
PS getting closer now - 12v regs / heatsinks almost ready to mount then I'll have a go with it. I need to make it a bit more safe and dress the wires properly - weekend maybe.
I'll fire some pics up soon
Andrew
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.