Below are the actual bill of materials, split in two for easier viewing:
Part 1:
Part 2:
I had to substitute some parts like one 3886 chip (shorted by mistake while troubleshooting and blew it up), the LM318 chip I'm using is some brand called JRC (I will use the TI opamp supplied as part of the kit), and I think I also replaced R1, R4 with silicon resistors as I could not get metal film types. But the most critical replacement was R11 which was actually faulty and gave so much birth pangs
I used a metal film similar to the one supplied in the kit.
Thanks.
Part 1:
Part 2:
I had to substitute some parts like one 3886 chip (shorted by mistake while troubleshooting and blew it up), the LM318 chip I'm using is some brand called JRC (I will use the TI opamp supplied as part of the kit), and I think I also replaced R1, R4 with silicon resistors as I could not get metal film types. But the most critical replacement was R11 which was actually faulty and gave so much birth pangs
I used a metal film similar to the one supplied in the kit.Thanks.
I don't know the different versions but I do remember that when playing with this lovely amp, a big improvement came from adding to the LM318 its own high quality regulated supply, meaning good Vregs...
If only I had planned a more spacious cabinet it would have been possible to add a small board for regulated supply with its own independent R-Core transformer.
And I guess I'll have to cut the traces on the boards for the opamp V+, Ground and V- power supply lines?
Or you could derive the power from the same supply (R->zeners?) and just add the Vregs in (vertical) daughterboards, as I did.
Cheers,
M.
Below are the actual bill of materials, split in two for easier viewing:
Part 1:
Part 2:
I had to substitute some parts like one 3886 chip (shorted by mistake while troubleshooting and blew it up), the LM318 chip I'm using is some brand called JRC (I will use the TI opamp supplied as part of the kit), and I think I also replaced R1, R4 with silicon resistors as I could not get metal film types. But the most critical replacement was R11 which was actually faulty and gave so much birth pangs
Thanks.
I think this is one of Siva's later Premium BOMs (post blackgate)
Maybe it can help
OK, that's a very interesting way to do it.
So did you have to break the board traces after the zeners, and then divert that to the vertical daughter board?
Did you use 3-pin reg chips to regulate? What would you recommend? I have used LM317/337 in the past.
OK, I swapped out the 2.2 uF to stranded 20 AWG wire (tinned copper). I think a thin veil has been lifted, making overall sound a bit cleaner and clearer, but the spectral balance remains as before.
I'd be happy to try regulated power to the LM318. To prove the concept, I think it will be possible to try powering the opamps from an external LM317/337 based reg without cutting any board traces.
Plain LM317/337 (+/-12V if I recall correctly) will do to prove the concept but I don't expect too much improvement (I may be wrong) soundwise. If you find that there is a worthwhile improvement, then you may think about better options...
As I like to complicate myself, I used TeddyRegs, which are way better sounding than the above. Now there's one even better, the SuperTeddyregs (wich may end being bigger) and many other Vregs that are "audiophile aproved", so to speak. But, it is a lot more work...
Good luck.
M.
As I like to complicate myself, I used TeddyRegs, which are way better sounding than the above. Now there's one even better, the SuperTeddyregs (wich may end being bigger) and many other Vregs that are "audiophile aproved", so to speak. But, it is a lot more work...
Good luck.
M.
Thanks for the good advice, M.
So I did the deed Temporarily I extended regulated supply from a DCB1 buffer to power just the opamps. Since the opamps are mounted on sockets, it was easy to bend pins 4 and 7 so that they don't insert into the socket, then a spare opamp socket was cut up to harvest the pins which were then used to connect the external DC supply. The DC supply is from LM317/337 with proper LC filtering, so not too shabby.
From the very first notes it was clear that the regulated supply changed the sound drastically. My main crib was the recessed highs. With the regulated supply to the opamps, the highs gained good extension. This led to improved resolution and improved image depth and image specificity. (Sorry for the "subjectivese").
But it also affected the midrange. Gone was the glorious midrange bloom. The midrange had flattened.
It also took away some of the midbass bump which gave the illusion of deeper bass reach and weight.
MyRef C was sounding more like a any other neutral amp. Since I already have plenty of that sound signature on both my big class A and TPA3118 class D, I think I'll stick to the tube-like flavour of MyRef Rev C. I can live without the ultimate detail or treble extension, but already missed the euphonic mids.
For good measure I've finally plugged in the TI LM318s. Will give it a good listen to see if these authentic opamps make a difference.
So I did the deed
Temporarily I extended regulated supply from a DCB1 buffer to power just the opamps. Since the opamps are mounted on sockets, it was easy to bend pins 4 and 7 so that they don't insert into the socket, then a spare opamp socket was cut up to harvest the pins which were then used to connect the external DC supply. The DC supply is from LM317/337 with proper LC filtering, so not too shabby. From the very first notes it was clear that the regulated supply changed the sound drastically. My main crib was the recessed highs. With the regulated supply to the opamps, the highs gained good extension. This led to improved resolution and improved image depth and image specificity. (Sorry for the "subjectivese").
But it also affected the midrange. Gone was the glorious midrange bloom. The midrange had flattened.
It also took away some of the midbass bump which gave the illusion of deeper bass reach and weight.
MyRef C was sounding more like a any other neutral amp. Since I already have plenty of that sound signature on both my big class A and TPA3118 class D, I think I'll stick to the tube-like flavour of MyRef Rev C. I can live without the ultimate detail or treble extension, but already missed the euphonic mids.
For good measure I've finally plugged in the TI LM318s. Will give it a good listen to see if these authentic opamps make a difference.
Topology and passives selection of the regulator can change the sound a lot.
When designing my My_Ref variation I've had best results with a simple regulated shunt power supply, see schematic:
https://drive.google.com/file/d/0B-qKRBuaKtzYRVQxeW1SNmNIYmc/view?usp=sharing
Thanks, Dario for your shuntreg circuit.
I'm currently powering Keantoken's Kuartlotron buffer and Hypnotoad low output MC phono preamp with this shuntreg: BluesPower Dual Shunt Regulator Power Supply Kit - Audio Kits
I've biased it for very high current (nearly 800 mA).
I'll try powering the LM318s with this shuntreg to see if the sound is more to my liking.
Besides component types, one more variable I observed while building the shuntreg I mentioned above is the change in sound with supply voltage. For example, the Kuartlotron buffer is designed to be powered by +/-10V. When I set the shuntreg to this voltage, I didn't hear anything special. I upped the voltage to about 14.3V and heard a completely different sound. It was very forward and in-your-face. It also played havoc with image size (bigger than would be normal). Aggressive would be the right word, though it did bring improved flow and cadence. More trials followed. 12V was rejected for losing the mojo of the flow, though the soundstage was more laidback and closer to my liking. Final voltage is 13.3V. It gives the best balance of musical flow while not being in your face, or projecting unrealistic image size.
Yesterday's experiment was at +/-10V only, since that's the supply for DCB1. In hindsight, I should have played around a bit with the voltage level.
I'm currently powering Keantoken's Kuartlotron buffer and Hypnotoad low output MC phono preamp with this shuntreg: BluesPower Dual Shunt Regulator Power Supply Kit - Audio Kits
I've biased it for very high current (nearly 800 mA).
I'll try powering the LM318s with this shuntreg to see if the sound is more to my liking.
Besides component types, one more variable I observed while building the shuntreg I mentioned above is the change in sound with supply voltage. For example, the Kuartlotron buffer is designed to be powered by +/-10V. When I set the shuntreg to this voltage, I didn't hear anything special. I upped the voltage to about 14.3V and heard a completely different sound. It was very forward and in-your-face. It also played havoc with image size (bigger than would be normal). Aggressive would be the right word, though it did bring improved flow and cadence. More trials followed. 12V was rejected for losing the mojo of the flow, though the soundstage was more laidback and closer to my liking. Final voltage is 13.3V. It gives the best balance of musical flow while not being in your face, or projecting unrealistic image size.
Yesterday's experiment was at +/-10V only, since that's the supply for DCB1. In hindsight, I should have played around a bit with the voltage level.
I am wondering if this reported sensitvity to opamp supply type is pointing to improper supply rail decoupling?
The opamp and the chipamp both need HF and MF supply rail decoupling.
The HF decoupling must be located at the chip pins and must be connected to a very close decoupling ground. If the route to the ground and back is long, then the HF portion of the decoupling cannot work.
The opamp and the chipamp both need HF and MF supply rail decoupling.
The HF decoupling must be located at the chip pins and must be connected to a very close decoupling ground. If the route to the ground and back is long, then the HF portion of the decoupling cannot work.
@Andrew: the opamps do have 100 nF on the +/- supply rails. I don't recall how close they're located to the opamp unless I physically check it. The LM3886 supply rails have 220uF and 100nF. Ditto about physical location.
I tried my shuntreg to power the opamps. Better subjective sound to my ears than the LM317/337. But in the end fatiguing, probably because the shuntreg voltage (+/-13.3V) is high for this application. Changing the voltage is not an option because I'll have to dismantle heat sink, board, cabling, etc to be able to swap out the zener. I think I'll let it rest here. I do like the sound of the MyRef, as is.
I tried my shuntreg to power the opamps. Better subjective sound to my ears than the LM317/337. But in the end fatiguing, probably because the shuntreg voltage (+/-13.3V) is high for this application. Changing the voltage is not an option because I'll have to dismantle heat sink, board, cabling, etc to be able to swap out the zener. I think I'll let it rest here. I do like the sound of the MyRef, as is.
Last edited:
Mine is the Linuxguru Rev C boards. Would you, by any chance, recall from memory if the decoupling caps are placed in close proximity to the opamp?
Also, for the LM317, do you feel adding to the existing 100 nF might help?