Hi,
I'm building a new GC with a pre stage (buf634 and OPA627 a la Carlos) and would like to include the low pass filter as suggested by Rassmusen.
Should I locate the LPF attached to the very chip? I'm planning the inverted Thorsten one, but because now will be buffered and the pot (10K) will be at the input of the OPA I don't know which the values for the filter will be.
Will be ok with the following values?
I'm building a new GC with a pre stage (buf634 and OPA627 a la Carlos) and would like to include the low pass filter as suggested by Rassmusen.
Should I locate the LPF attached to the very chip? I'm planning the inverted Thorsten one, but because now will be buffered and the pot (10K) will be at the input of the OPA I don't know which the values for the filter will be.
Will be ok with the following values?
Attachments
The BUF634 makes sense on a standalone pre.
You can drive your power amps meters away from the pre with no problems.
Raka, the filter should be inside your GC, not the pre.
But if you're going to put everyting inside the same box, the pre doesn't make sense anymore, just use the OPA627s with the filter, as on Nuuk's thread.
You can drive your power amps meters away from the pre with no problems.
Raka, the filter should be inside your GC, not the pre.
But if you're going to put everyting inside the same box, the pre doesn't make sense anymore, just use the OPA627s with the filter, as on Nuuk's thread.
Inverting or not?
Hi guys,
there is something that i do not quite follow here:
The desition to build an inverter GC is for the sonic benefits of the inverter configuration, right?
But the problem associated with this config is low input impedance, right?
So then you compensate the low impendance with a buffer. But the buffer is NON INVERTNG!.So say good bye to the advantage of the inverting GC config.
So are we going on circles here or what?
Wouldn't be more practical to just build the GC as original (non inverting) if you need the high input impedance?
To put a filter on the non inverting input would be may be easier.
I say this because I build 9 different GC with inverting and non inv, config and the sound in not bad at all in either config.
cheers
Ric
Hi guys,
there is something that i do not quite follow here:
The desition to build an inverter GC is for the sonic benefits of the inverter configuration, right?
But the problem associated with this config is low input impedance, right?
So then you compensate the low impendance with a buffer. But the buffer is NON INVERTNG!.So say good bye to the advantage of the inverting GC config.
So are we going on circles here or what?
Wouldn't be more practical to just build the GC as original (non inverting) if you need the high input impedance?
To put a filter on the non inverting input would be may be easier.
I say this because I build 9 different GC with inverting and non inv, config and the sound in not bad at all in either config.
cheers
Ric
You see, that's something I also don't understand. If you need a filter on the GC, it means that something went wrong when building it. The whole advantage of that amp is simplicity and minimized number of components, which in the end produce spectacular results. I've built few of those GC and it never crossed my mind to add a filter because highs were no good, actually they are better than in all other amps I tried so far.
When I hear people commenting that the buffer improves the sound, I take it with a big dose of sceptisism, because I'm still afraid that the cure might be worse than the disease. That's also one reason I still didn't try a buffer
That's also a reason that I shouldn't be really commenting on that either
When I hear people commenting that the buffer improves the sound, I take it with a big dose of sceptisism, because I'm still afraid that the cure might be worse than the disease. That's also one reason I still didn't try a buffer
That's also a reason that I shouldn't be really commenting on that either
I also built few GCs (5-6) in inv. and noninv. configurations and to my ears there is much greater difference in sound between various IC's than between inv. and noninv. configuration.
Also, I tried JFET buffered lowpass in front of inverted LM3875 GC and it really improves sound, but moamps' tubed version is even better.
Also, I tried JFET buffered lowpass in front of inverted LM3875 GC and it really improves sound, but moamps' tubed version is even better.
Peter Daniel said:
Hi,
I don't think so. Something is wrong with chip itself. Chip isn't ideal. Pease read
http://www.national.com/ms/LB/LB-19.pdf (only two pages)
There you can find good explanation of slew rate limiting.
Regards
First may I suggest that you read and understand the signature line to this post.
Sometimes we seem to get into ruts on this forum where a topic is raised, dealt with and then brought up in another thread, may be in the hopes that a different answer will be arrived at!
I don't know how many have actually tried the buffered version of the IGC but I do know what I hear and we are not talking a minor difference, we are talking about a big improvement. This week a friend came round and listened to my buffered ICG, having listened to the same amp but unbuffered last week. He immediately noticed the improvement for himself.
The recurring theme that I talk about is this business of putting forward a theory to say that something shouldn't sound better, therefore it won't.
When I came to this forum originally and started reading about chip amps, I thought to myself that they couldn't possibly sound as good as the reports. BUT I didn't start posting to that effect because I hadn't built one for myself, or heard one that somebody else had built.
I don't want to pick on anybody in particular but I have to say that I am surprised at one contributor to this thread who has generally put his mouth where his soldering iron is and spoken from experience rather than conjecture, making a statement about the buffered ICG instead of trying it for himself!
End of lecture
Sometimes we seem to get into ruts on this forum where a topic is raised, dealt with and then brought up in another thread, may be in the hopes that a different answer will be arrived at!
I don't know how many have actually tried the buffered version of the IGC but I do know what I hear and we are not talking a minor difference, we are talking about a big improvement. This week a friend came round and listened to my buffered ICG, having listened to the same amp but unbuffered last week. He immediately noticed the improvement for himself.
The recurring theme that I talk about is this business of putting forward a theory to say that something shouldn't sound better, therefore it won't.
When I came to this forum originally and started reading about chip amps, I thought to myself that they couldn't possibly sound as good as the reports. BUT I didn't start posting to that effect because I hadn't built one for myself, or heard one that somebody else had built.
I don't want to pick on anybody in particular but I have to say that I am surprised at one contributor to this thread who has generally put his mouth where his soldering iron is and spoken from experience rather than conjecture, making a statement about the buffered ICG instead of trying it for himself!
End of lecture
Raka, if you have an OPA627, and you have my circuit diagram, just go for it as it is.
Listen to it and if you think that it needs improving, come back and ask advice, or just try variations on the circuit as you wish.
I messed around with that circuit and spent time getting it to work 'properly'. I then took the time to write up all the work on Decibel Dungeon to save others who wanted to try this idea repeating all that work. Joe R and Pedja also did a lot of work to pioneer the concept and make their knowledge available to others.
I have always said that planning is at least 80% of (successful) DIY hi-fi. But perfecting is done, not so much in the planning stage but after the building stage when you have something to improve!
My old guru told us once 'A woman always does the washing up four times, once in the sink and three times in her head'.
'Many a true word is spoken in jest'
Listen to it and if you think that it needs improving, come back and ask advice, or just try variations on the circuit as you wish.
I messed around with that circuit and spent time getting it to work 'properly'. I then took the time to write up all the work on Decibel Dungeon to save others who wanted to try this idea repeating all that work. Joe R and Pedja also did a lot of work to pioneer the concept and make their knowledge available to others.
I have always said that planning is at least 80% of (successful) DIY hi-fi. But perfecting is done, not so much in the planning stage but after the building stage when you have something to improve!
My old guru told us once 'A woman always does the washing up four times, once in the sink and three times in her head'.
'Many a true word is spoken in jest'
Hi Nuuk ,
Thanks for your advice. I wanted to use the BUF634 but I also want an integrated amp, so I'll use your scheme without it and just the opa627.
Just a question, in the diagram I copied from your site, the 10K resistor to the right is part of the "standard Thorsten" GC, isn't it? So I only have to add the parts surrounded and change the input pot for a 100K fixed resistor, right?
Would be any problem with the turn on-off?
Thanks for your advice. I wanted to use the BUF634 but I also want an integrated amp, so I'll use your scheme without it and just the opa627.
Just a question, in the diagram I copied from your site, the 10K resistor to the right is part of the "standard Thorsten" GC, isn't it? So I only have to add the parts surrounded and change the input pot for a 100K fixed resistor, right?
Would be any problem with the turn on-off?
Re: Inverting or not?
It is the good question, but two things might be important why this works. First, OPA627 surely works better as the input device considering slew induced distortion (and after it you have a filter), and second, I think that the originator of the “invert da sukah” theory would say the better the opamp – the less difference between the inverting and the non-inverting configuration (neglecting the power capabilities and compared directly, LM3875 doesn’t have any chance against OPA627).
Pedja
Hi Ric,Ricren said:
It is the good question, but two things might be important why this works. First, OPA627 surely works better as the input device considering slew induced distortion (and after it you have a filter), and second, I think that the originator of the “invert da sukah” theory would say the better the opamp – the less difference between the inverting and the non-inverting configuration (neglecting the power capabilities and compared directly, LM3875 doesn’t have any chance against OPA627).
Pedja
Nuuk said:
I don't want to pick on anybody in particular but I have to say that I am surprised at one contributor to this thread who has generally put his mouth where his soldering iron is and spoken from experience rather than conjecture, making a statement about the buffered ICG instead of trying it for himself!
I'm just discouraged by previous suggestion which I tried. I built amp based on TDA7294, which supposed to be the best GC chip, yet it appeares to be the worse. I built amp based on OPA549, still didn't like it compared to LM chip. I tried biasing the output with resistor from raild, which at one time was very popular trick, it also screwed up the sound of the amp (although some guys went into so much trouble as inventing active circuits for that purpose).
So all those "supposedly" improvements didn't work out. I have a OPA627 at the output of my Marantz CD player, and although TDA1541 suppose to be better than TDA1543, the 1543 chip with passive output sounds much better (even in a separate DAC) than the DAC with OPA627. I tried many tweaks to improve the circuit, yet the loss of resolution is clearly noticable.
But, since this is a popular subject these days, I'll built the buffer today and see what it's all about. I should've done it long ago
Nuuk, I didn't see you mentioned what PS you use with the buffer?
Hi Peter,
I built GCs with TDA1514, TDA7294, LM1875, LM3875 and LM3886 some inverted and some not and some in both versions.
I never used "audiophile approved" caps and resistors, but all were built with same, standard-quality components in extra tight layout. I still prefere TDA7294! I didn't like LM3875 inverted or not, but moamps' buffered version sounded wonderful, so I tried JFET buffer which helped but didn't sound as good as tubes.
I also built several nonos DACs; TDA1541, TDA1543 and, very recently, AD1865. To my ears TDA1541 is the best!
We are probably listening to different kind of music too, and like different food .
So... different people, different tastes, but you really shouldn't make statements about something you haven't heard.
Having for years been in the mfg. end of electrical, electro-mech. and mech, assys. and worked in QC/QA i can assure everyone that there are no 2 exactly performing of anything.Everything is built to a set of tolerences. If a test subjects performance falls within the parameters of the tolerences then its accepted.Also there is very little if any 100% testing of any product line due to SPC (stastical process control).If a buffer (or variable resistor or cap) helps bring 2 or more performances of a chip into more exacting results then its necessary.
ron
BTW. Interesting reading (if you are a geek like me) about Demming
and his try to impliment SPC in america.America didnt listen and japan did ,with great interest.This helped bring about the improved (over america) product quality of japan and established them in just about every market.
ron
BTW. Interesting reading (if you are a geek like me) about Demming
and his try to impliment SPC in america.America didnt listen and japan did ,with great interest.This helped bring about the improved (over america) product quality of japan and established them in just about every market.
Hi Raka,
Yes, the 10K is the 10K input resistor of the IGC circuit.
With the buffer in place, almost any pot/attenuator value will work, ie it doesn't have to be 100K, that was what I had handy.
As regards problems turning the power off, this is a tricky one. The first PSU that I tried with the buffer circuit did produce a high offset (a few volts albeit momentarily) when the power was switched off. You wouldn't (or shouldn't) want that going through your ICG circuit!
I tired another PSU and only get mVs now when the PSU for the buffer is switched off. However, I always turn off the amps PSU FIRST before I turn off the PSU for the buffer.
As regards building a preamp, having the OPA627 buffer before the ICG is just about what a preamp would consist of! Believe me, if yours sounds like mine does, you won't care whether you call it a buffer or a pre!
Peter D, I described the PSU in my original thread
but here it is again.
--------------------------------------------------------------------------
The PSU for the buffer was another 'already have' item and I didn't have time yesterday evening to write up full details but will later.
It is a fairly standard job, 18-0-18 torroid, rectifer bridge with Schottkeys if I remember correctly, 2 x 4700uF with smaller bypass caps and one stage of regulation (317/337) in one box, then another stage of regulation (317/337) for each channel closer to the buffers.
-----------------------------------------------------------------------------
I am working on a circuit diagram to go on Decibel Dungeon, possibly this weekend.
but here it is again.
--------------------------------------------------------------------------
The PSU for the buffer was another 'already have' item and I didn't have time yesterday evening to write up full details but will later.
It is a fairly standard job, 18-0-18 torroid, rectifer bridge with Schottkeys if I remember correctly, 2 x 4700uF with smaller bypass caps and one stage of regulation (317/337) in one box, then another stage of regulation (317/337) for each channel closer to the buffers.
-----------------------------------------------------------------------------
I am working on a circuit diagram to go on Decibel Dungeon, possibly this weekend.
By a strange coincidence, I had just posted the previous message and we had a short power cut here (a few seconds).
I left my PC to reboot and went upstairs where the buffered IGC is to check if removing the power from the buffer while the GC was powered up too had done any damage but I'm pleased to say that it is fine and so are the speakers!
I left my PC to reboot and went upstairs where the buffered IGC is to check if removing the power from the buffer while the GC was powered up too had done any damage but I'm pleased to say that it is fine and so are the speakers!
vuki said:
I don't see myself doing statement about anything. I just expressed my opinion that it's hard for me to believe that GC with buffer might sound better. And I'm listening to all sort of music and not only myself I decide about parts and topology choices. There are at least two more people involved, with probably better ears than mine.
If we talking about making statements, I can also accuse you of doing that. It was actually your influence that I decided to try TDA chip, as you claimed was the best. Well, I tried it and it seems to be the worst. I used exactly same chassis and major parts as my LM version and there is simply no comparison. If TDA7294 is the best, why everybody's using LM 3875 chips? BTW, I wasted at least 3 days fidddling with this dissapointing chip.
As to 1541 and 1543 DAC comparison, I don't think I'm the only one claiming that 1543 might actually be better.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.