Hi Lazyfly.
you will always have those issues with this particular design but it is only when you play very load.
I am checking a few things , but i think that I might have over bias the Lm318 and that is causing the clipping , just a thought. will check it during this week (not that i will remove it if it is a problem, because it makes the amp focus much better)
this is like a race car. to get the max out of it you will have to sacrifice something else
you will always have those issues with this particular design but it is only when you play very load.
I am checking a few things , but i think that I might have over bias the Lm318 and that is causing the clipping , just a thought. will check it during this week (not that i will remove it if it is a problem, because it makes the amp focus much better)
this is like a race car. to get the max out of it you will have to sacrifice something else
And the winner is....
Rudi and I tested my amp (Pre=AD815 as per CFM, power=NI LM3875, MUR860’s, 300VA per channel, 1000uF per PS leg) against the Excalibur.
Before we get into the detail, Rudi hade made a mod to his DAC (reducing the output cap from 1.5uF to 1.0uF) which I think improved the overall sound of his system-basically the upper bass was not as warm, which I preferred.
Onto the shoot-out:
My amp had highs that were comparable to the X-amp, detailed, extended. This reflects the benefit of the AD815, since on its own, the NIGC 3875 highs tended to be relatively harsh.
That’s were the similarities end unfortunately.
The X-amp had better resolution and separation and presence on the mids. In comparison, my amp sounded a bit congested in the mids.
Also notable was the fact that the X-amp had much better layering, so that the various instruments existed more in their own space rather than being melded into one. But at the same time the amp still sounded organic, i.e. musical.
Bass performance was interesting: My amp went deep (the benefit of the AD815), but not as deep and fundamental as the X-amp-it is a real heavy weight in this department, some of the bass you feel more than hear. Also I’d say that the X-amp was perhaps more detailed.
Overall, the X-amp was more coherent/clean/clear and ultimately musical and enjoyable, especially in the mids vs my own. Less hifi, more music!
I would say that my own 3875 was transformed by the use of the AD815 and Rudi and I will be exploring using it on the X-amp.
Will keep ya’ll posted!
Ryan
Rudi and I tested my amp (Pre=AD815 as per CFM, power=NI LM3875, MUR860’s, 300VA per channel, 1000uF per PS leg) against the Excalibur.
Before we get into the detail, Rudi hade made a mod to his DAC (reducing the output cap from 1.5uF to 1.0uF) which I think improved the overall sound of his system-basically the upper bass was not as warm, which I preferred.
Onto the shoot-out:
My amp had highs that were comparable to the X-amp, detailed, extended. This reflects the benefit of the AD815, since on its own, the NIGC 3875 highs tended to be relatively harsh.
That’s were the similarities end unfortunately.
The X-amp had better resolution and separation and presence on the mids. In comparison, my amp sounded a bit congested in the mids.
Also notable was the fact that the X-amp had much better layering, so that the various instruments existed more in their own space rather than being melded into one. But at the same time the amp still sounded organic, i.e. musical.
Bass performance was interesting: My amp went deep (the benefit of the AD815), but not as deep and fundamental as the X-amp-it is a real heavy weight in this department, some of the bass you feel more than hear. Also I’d say that the X-amp was perhaps more detailed.
Overall, the X-amp was more coherent/clean/clear and ultimately musical and enjoyable, especially in the mids vs my own. Less hifi, more music!
I would say that my own 3875 was transformed by the use of the AD815 and Rudi and I will be exploring using it on the X-amp.
Will keep ya’ll posted!
Ryan
This has been one of those shoot-outs that had me a bit nervous.
I have heard allot of the Lm3875 and the combination of this with an AD815. To top it Ryan had a DACT CT2 to control all of this
My combination were a passive ALPS and the X-Calibre as is published
The X-Calibre is much more airy and much less hash, much better definition and also allot quieter. the LM3875 sounded closed in and the mids were lifeless. The top were the only part of the amp that sounded realistic.
The LM3875 sounded very much like the stock standard Ref C that i have built (including the 550va trafo's and Sprague caps which transformed the amp) I do not think the commercial Ref-C amp can sound better than the Ref-C I have implemented.
It will be interesting to hear the difference between the standard TP Ref C and the X-Calibre
I have heard allot of the Lm3875 and the combination of this with an AD815. To top it Ryan had a DACT CT2 to control all of this
My combination were a passive ALPS and the X-Calibre as is published
The X-Calibre is much more airy and much less hash, much better definition and also allot quieter. the LM3875 sounded closed in and the mids were lifeless. The top were the only part of the amp that sounded realistic.
The LM3875 sounded very much like the stock standard Ref C that i have built (including the 550va trafo's and Sprague caps which transformed the amp) I do not think the commercial Ref-C amp can sound better than the Ref-C I have implemented.
It will be interesting to hear the difference between the standard TP Ref C and the X-Calibre
I finally just fitted the Carlos snubber to this amp (0.47R-1n5) across the LM3886 pins as well as keeping the 100nf
Sorry guys but the snubber does improve this circuit IMO, if you can't hear the difference you must me deaf
I tried the MUR860's again but I still prefer the MBR10100
I also fitted the 0.47R-47uf snubber on the outputs of my regulator boards which power the LM318's, the sound has improved some more but not as noticeable as the ones for the LM3886 which is probably because of it helping those big 10k caps
A simple and cheap mod, worth trying at least
Sorry guys but the snubber does improve this circuit IMO, if you can't hear the difference you must me deaf
I tried the MUR860's again but I still prefer the MBR10100
I also fitted the 0.47R-47uf snubber on the outputs of my regulator boards which power the LM318's, the sound has improved some more but not as noticeable as the ones for the LM3886 which is probably because of it helping those big 10k caps
A simple and cheap mod, worth trying at least
BrianDonegan said:
Hi Brian. I have done some simulations and the simulations did not show any ill effect with the changed voltages. When I checked it with a scope I also did no find any strange behaviour.
By changing small values on the compensation network is far worse. You will make it unstable in an instant. That is part of the complexity.
Hi Rudi,
I asked about the clipping because I think I may have crossed its path yesterday. I'll do a little testing today - bypass the pre (which is set to unity anyway) and run passive. I'm using a 50k pot though which may affect this behaviour.
It appears quite a bit earlier (in relation to pot position) than the amp it's replaced which is Brian Bell's 3886 circuit based on Carlos' snubbreed PS. While it could be easy to dismiss the 'volume control level' method as inaccurate (it is) this amp does not go as loud using the same cdp, speakers, cabling etc. Speakers are Peerless 2 way ~89dB, PMt's designed by Rabbitz.
I'll get back to you with some results a little later.
It pains me to say this Rudi but I feel your transformers VA rating is a little low... BAHAHAHA... kidding
I asked about the clipping because I think I may have crossed its path yesterday. I'll do a little testing today - bypass the pre (which is set to unity anyway) and run passive. I'm using a 50k pot though which may affect this behaviour.
It appears quite a bit earlier (in relation to pot position) than the amp it's replaced which is Brian Bell's 3886 circuit based on Carlos' snubbreed PS. While it could be easy to dismiss the 'volume control level' method as inaccurate (it is) this amp does not go as loud using the same cdp, speakers, cabling etc. Speakers are Peerless 2 way ~89dB, PMt's designed by Rabbitz.
I'll get back to you with some results a little later.
It pains me to say this Rudi but I feel your transformers VA rating is a little low... BAHAHAHA... kidding
Re: And the winner is....
Trying to solve jitter and/or analog stage problems by reducing the output coupling caps?
That's the easy solution, but far from perfect.
Ryan, you are comparing the uncomparable.
Please leave once and for all that low capacitance chipamp. What you are hearing is the amp sweating to drive those speakers. It can sound as soubtle as that, until you hear the REAL THING.
Check the zip file I posted a couple of pages ago.
Follow that strictly and make a new amp, or modify your amp. I prefer the LM3886, anyway, but you will have very good results with the LM3875.
You will wonder from where all that power comes from.
You will also wonder why have you once thought about paralleling chips (which degrades the sound).
And what you will listen, in all aspects, will amaze you.
For unregulated PSU I recommend 20,000uF minimum after the series resistors. I'm using around 30,000uF.
And 2,200uF caps ON the chips, snubberized.
Notice that the amp on this thread is starting to have some similarities to what I've been recommending for the PSU.
Another reason why it also uncomparable the two amps that you listened to?
Because on your amp you are driving the LM3875 chip with a pot or the AD815 pre.
On Mauro's amp (or on the 'X-calibre') you are driving an LM318 op-amp.
You have already built my preamp.
As I know you also are a 'hardcore' diyer I'm challenging you to build my amp, strictly following the schematics, as you did for my pre.
Dr.H said:
Trying to solve jitter and/or analog stage problems by reducing the output coupling caps?
That's the easy solution, but far from perfect.
Dr.H said:
Ryan, you are comparing the uncomparable.
Please leave once and for all that low capacitance chipamp. What you are hearing is the amp sweating to drive those speakers. It can sound as soubtle as that, until you hear the REAL THING.
Check the zip file I posted a couple of pages ago.
Follow that strictly and make a new amp, or modify your amp. I prefer the LM3886, anyway, but you will have very good results with the LM3875.
You will wonder from where all that power comes from.
You will also wonder why have you once thought about paralleling chips (which degrades the sound).
And what you will listen, in all aspects, will amaze you.
For unregulated PSU I recommend 20,000uF minimum after the series resistors. I'm using around 30,000uF.
And 2,200uF caps ON the chips, snubberized.
Notice that the amp on this thread is starting to have some similarities to what I've been recommending for the PSU.
Another reason why it also uncomparable the two amps that you listened to?
Because on your amp you are driving the LM3875 chip with a pot or the AD815 pre.
On Mauro's amp (or on the 'X-calibre') you are driving an LM318 op-amp.
You have already built my preamp.
As I know you also are a 'hardcore' diyer
I'm challenging you to build my amp, strictly following the schematics, as you did for my pre.t. said:Sorry guys but the snubber does improve this circuit IMO, if you can't hear the difference you must me deaf
I will not comment, this time.
t. said:
It's curious you say that, because I have MBR16100 schottky diodes (3-leg TO-220) on my regulated PSU and I like them very much, but I never tried them on an unregulated PSU.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.