My_Ref C
It does not just add another chip. The entire power supply section is doubled. Twice as many rectifiers and filter caps.
From talking to Mauro, it took a lot of interactions in the compensation network to get stability and fast settling time for both single and dual chip builds in the Evo. The previous versions were unstable when two chips were paralleled.
He also was not impressed with the TP mono boards. Felt the layouts and ground plane design was not as good as the original My_Ref stereo boards. The comment was the current nets were not optimized.
The original boards were high quality. Even had a gold mask layer.
The design criteria was not just a composite design with a current pump exercise. Many aspects were optimized to allow fast recovery from overload, quick settling times, along with low THD, IMD, and linearity.
It is fantastic that Clave and George have continued the development. The measurements I saw that George ran are fantastic. Great numbers.
The dual LM3886 helps with low impedance speakers. The power chip has limited current output. If the speakers impedance drops below 4 ohms the Full Evo is best. If the speakers are 5 ohms and higher a single chip version should be better sonically.
But George’s tests and my personal experience do not support the single chip superiority. I have a Full Evo and it drove B&W 801 Matrix 2’s very well. Great bass control. Also it sounded great with 15 ohm horns that were close to 100 dB efficient. It handled current hungry speakers and very easy loads equally well.
I have a C version in the attic. It is the Twisted Pear Mono hoards. It does not sound as clear as the Evo. Also only the Evo is designed for parallel outputs. One item that makes the Full Evo better for the Full version is the power supply.Apologies for the late reply.
Thanks a lot Naaling for sharing for the PCB layout. This is something which I can get it done without much complexities. .
I got few more queries like
1. Is it possible to parallel/bridge MyRef version to increase output power?
2. Is there any matching pre-amp circuit available for MyRef?
Once again thank you all for your time
-Sha
It does not just add another chip. The entire power supply section is doubled. Twice as many rectifiers and filter caps.
From talking to Mauro, it took a lot of interactions in the compensation network to get stability and fast settling time for both single and dual chip builds in the Evo. The previous versions were unstable when two chips were paralleled.
He also was not impressed with the TP mono boards. Felt the layouts and ground plane design was not as good as the original My_Ref stereo boards. The comment was the current nets were not optimized.
The original boards were high quality. Even had a gold mask layer.
The design criteria was not just a composite design with a current pump exercise. Many aspects were optimized to allow fast recovery from overload, quick settling times, along with low THD, IMD, and linearity.
It is fantastic that Clave and George have continued the development. The measurements I saw that George ran are fantastic. Great numbers.
The dual LM3886 helps with low impedance speakers. The power chip has limited current output. If the speakers impedance drops below 4 ohms the Full Evo is best. If the speakers are 5 ohms and higher a single chip version should be better sonically.
But George’s tests and my personal experience do not support the single chip superiority. I have a Full Evo and it drove B&W 801 Matrix 2’s very well. Great bass control. Also it sounded great with 15 ohm horns that were close to 100 dB efficient. It handled current hungry speakers and very easy loads equally well.
Panelhead
A single LM3886 at 25V rails will drive 3.5R loads fairly well. Depends on the chip you got, some will dump ~8A but most can dump 5A. Yes, the TwistedPear board had a poor layout for this type of amp and that limited its potential but if yours is single sided, it'll make it easier to fix the ground plane, no? Would you be interested in revisiting this board to get the most out of it in memory of Mauro? JosephK has modeled some of Dario's boards so he has most of the models. If he decides to participate here, we can add the circuit trace parasitics to them so as to optimize what we have. We may not be able to achieve the 10x reduction in distortion of their effort but I suspect we can get it <100ppm for every measure.
A single LM3886 at 25V rails will drive 3.5R loads fairly well. Depends on the chip you got, some will dump ~8A but most can dump 5A. Yes, the TwistedPear board had a poor layout for this type of amp and that limited its potential but if yours is single sided, it'll make it easier to fix the ground plane, no? Would you be interested in revisiting this board to get the most out of it in memory of Mauro? JosephK has modeled some of Dario's boards so he has most of the models. If he decides to participate here, we can add the circuit trace parasitics to them so as to optimize what we have. We may not be able to achieve the 10x reduction in distortion of their effort but I suspect we can get it <100ppm for every measure.
Last edited:
Where has this suddenly come from?
Since I decided to build a My_Ref RevC about 4 years ago, I think I have read just about every post in every thread on this topic, and this is news to me.
Anyone who has followed the the design of TP boards will know that Russ designed the boards on the open forum and consulted directly with Mauro at every stage of the design. If Mauro had been unhappy with the layout at that time, then he would have told Russ and and the layout would have been changed.
P.S.
If you have an LM3886 that can only "dump" 5A, then you have a fake chip!
Last edited:
Naaling
Yes, Mauro did participate in the layout and gave his nod. I cannot say if he knew all the while that the layout was sub-optimal, suspected it might be or learned later that it was. For my own part, I was unaware how important the layout becomes when the parts are operating in the MHz range even if the bandwidth involved is nowhere near that high. Even things as seemingly trivial as cap recharging become sonic issues. For example, in the TP layout, what is between the signal ground and the power ground?
Yes, Mauro did participate in the layout and gave his nod. I cannot say if he knew all the while that the layout was sub-optimal, suspected it might be or learned later that it was. For my own part, I was unaware how important the layout becomes when the parts are operating in the MHz range even if the bandwidth involved is nowhere near that high. Even things as seemingly trivial as cap recharging become sonic issues. For example, in the TP layout, what is between the signal ground and the power ground?
Hi there
I have this Mauro amp design, I suppose it is rev.c. Wanted to verify if any one ever built this amp.
Also I am having trouble finding couple of components for this amp. Kindly guide if mentioned alternatives will work.
Element14::Fides Electronics
6 Amp Bridge Rectifier - BR68
BC639 NPN High Current Transistor TO-92 Package buy online at Low Price in India - ElectronicsComp.com
BC556 PNP General Purpose Transistor TO-92 Package buy online at Low Price in India - ElectronicsComp.com
Also I am not getting National LM318. I am getting only LM318N/NOPB, LM318P both Texas Instruments. will this work.
Apart from that I am seeing a huge variance in price of 1N4001 and 12v Zener diodes. Will the cheaper ones work.
Kindly Guide
Regards
Vikram
I have this Mauro amp design, I suppose it is rev.c. Wanted to verify if any one ever built this amp.
Also I am having trouble finding couple of components for this amp. Kindly guide if mentioned alternatives will work.
Element14::Fides Electronics
6 Amp Bridge Rectifier - BR68
BC639 NPN High Current Transistor TO-92 Package buy online at Low Price in India - ElectronicsComp.com
BC556 PNP General Purpose Transistor TO-92 Package buy online at Low Price in India - ElectronicsComp.com
Also I am not getting National LM318. I am getting only LM318N/NOPB, LM318P both Texas Instruments. will this work.
Apart from that I am seeing a huge variance in price of 1N4001 and 12v Zener diodes. Will the cheaper ones work.
Kindly Guide
Regards
Vikram
Last edited: