Guys, two years ago I've just made this amplifier and was working very fine.
Then, I asked a friend to re-design the pcb and my friend have designed a double side pcb. With this new pcb the amplifier had a better sound, it was quitere. I've tested the amplifier with different speaker cables and with different speakers. It worked very good with Epos M12.2, Epos M5 and ProAc Studio110. As speaker cables it worked very good with DNM cables, ProAc speaker cables and some special Italian cables.
I've compared this amplifier 1 to 1 with Creek 5350SE and with Atoll IN100. The Creek was warmer and relaxed but the soundstage was not as deep as with my amplifier. The Atoll IN100 was also warmer but not as Creek. Al in one, this amplifier is a very good one with a clean and clear sound. A very good value.
This amplifier works better with speakers wich are an ease load.
In time I've made 3 amplifiers for my friends and all of them are very satisfied.
Last year I've tried and I've made for a friend an amplifier just with LM3886 , a real gainclone, but this is a diferent story.
Then, I asked a friend to re-design the pcb and my friend have designed a double side pcb. With this new pcb the amplifier had a better sound, it was quitere. I've tested the amplifier with different speaker cables and with different speakers. It worked very good with Epos M12.2, Epos M5 and ProAc Studio110. As speaker cables it worked very good with DNM cables, ProAc speaker cables and some special Italian cables.
I've compared this amplifier 1 to 1 with Creek 5350SE and with Atoll IN100. The Creek was warmer and relaxed but the soundstage was not as deep as with my amplifier. The Atoll IN100 was also warmer but not as Creek. Al in one, this amplifier is a very good one with a clean and clear sound. A very good value.
This amplifier works better with speakers wich are an ease load.
In time I've made 3 amplifiers for my friends and all of them are very satisfied.
Last year I've tried and I've made for a friend an amplifier just with LM3886 , a real gainclone, but this is a diferent story.
Yes, indeed, it is a very good amp. Yet, I found that, as maybe with all circuits, it improves significantly with good regulated supply to the lm318 opamp (did I remember well the number?). I suspect that it will also improve with a good regulated power, but now I am in the middle of many projects and I can't test this hypothesis. 🙂
Cheers,
M.
Cheers,
M.
I'm interested in building this, but would also like to know what active DIY pre amp would be recommended to go with it. I've been looking at (dare I mention his name here) Carlos's AD815 pre amp, but I picked up that this might not be the best choice for the amp. Not sure why though.
Suggestions? (I'm can make my own PCB's from schematics (within reason), but a kit or pre built board would be ok).
Suggestions? (I'm can make my own PCB's from schematics (within reason), but a kit or pre built board would be ok).
Sorry, you said the name. Cant help ya 🙂
Do you have a link to it? Or do know the output impedance?
Uriah
Do you have a link to it? Or do know the output impedance?
Uriah
I have made a few pcbs and a few years ago tried to etch the single sided version. I dont know what went wrong but it never turned on. So I salvaged my parts and stuffed them in a regular PCB on the last group buy.
Uriah
Uriah
No, until I broke mine in a move I was running it with a Lightspeed which is totally passive to the signal and is absolutely mind blowingly good.
This also means you could use a switched pot or any type pot or other type passive attenuator.
Uriah
This also means you could use a switched pot or any type pot or other type passive attenuator.
Uriah
I do - my power amp will be about 10 feet from my source - mp3 player and CD player. MP3 I'm not worried about, but the distance between the CD Player and the amp will require a pre amp.
no. The maximum acceptable distance is determined by the source impedance and the ability of the source to drive current into a capacitive load.
You need gain if the voltages are mismatched, such that the source cannot drive the receiver to adequate SPL/volume.
Could someone please tell me how this configuration sounds compared to Briantgt's with a discrete class a buffer and what the main changes are other than the opamp feedback setup and speaker protection extras?
I started building brian's a year ago and never finished and now am wondering if it's worth abandoning it in favour or this or is the performance the same with minus the incorporated opamp?
I'd really appreciate comments I've been busy and havent had time to keep track of this.
Thanks
I started building brian's a year ago and never finished and now am wondering if it's worth abandoning it in favour or this or is the performance the same with minus the incorporated opamp?
I'd really appreciate comments I've been busy and havent had time to keep track of this.
Thanks
I have built several of BrianGT's LM3886 boards. I love them. So simple so cheap so nice to listen to. I especially like to use his Mini A power supply with the chipamp boards.
The MyRef does blow it away. I used the B1 with the MyRef and I also used my Lightspeed with the MyRef. Both were good, using the Lightspeed is incredible.
This is a Howland Current Pump. Its been almost exactly a year since I listened to the MyRef. I broke it in our move to Texas. Will be building it again in less than a month. I have had the BrianGT going since the MyRef broke so I will be able to directly compare. cant wait to build it again. It really is the top of the line using LM3886.
That Howland Current Pump is the big difference.
Uriah
The MyRef does blow it away. I used the B1 with the MyRef and I also used my Lightspeed with the MyRef. Both were good, using the Lightspeed is incredible.
This is a Howland Current Pump. Its been almost exactly a year since I listened to the MyRef. I broke it in our move to Texas. Will be building it again in less than a month. I have had the BrianGT going since the MyRef broke so I will be able to directly compare. cant wait to build it again. It really is the top of the line using LM3886.
That Howland Current Pump is the big difference.
Uriah
I will be feeding my MyRef_C with a Salas DCB1 and aLightspeed attenuator. The DCB1 is amongst the best pre's that exist now. Is active but has no gain, so it acts as a buffer, matching impedances to avoid frecuency rolloffs.
Uh? +-35Vdc is about the supply rails... the actual max output voltage is somewhere around 60Vpp. But regiregi22 asked about input voltage for max output, that is sensibility. 🙂
Gain is about 30dB, less than 1Vrms input is enough for full power. Thus, with modern sources the my_ref does not really need any extra gain. But it likes a lot (almost "requires" to) being driven by a low and constant impedance, so for best performance a low output-Z active preamp or buffer is highly recommended.
Gain is about 30dB, less than 1Vrms input is enough for full power. Thus, with modern sources the my_ref does not really need any extra gain. But it likes a lot (almost "requires" to) being driven by a low and constant impedance, so for best performance a low output-Z active preamp or buffer is highly recommended.
In fact, I didn't, It was Matt 😛 😉
BTW, I have read in this thread that gain is not 30dB, it it 30x. 1v in gives you 30v out. Correct me if I am wrong, please.
Your full output voltage will be defined by the clipping point of the amplifier with a given load. You will have different clipping points depending if using 4 or 8 Ohm, so sensibility will be different.
To measure, you have to attach a high wattage load resistor, and looking at the scope, you up the input voltage until the voltage out clips. That's your sensibility input voltage.
The gain is about 30dB. Incidentally, 20 * Log(30) ~= 30 dB
And 1 Vrms ~= 2.82 Vpp x 30 = 84.6 Vpp out, much more than what the my_ref can do.
Current limit is (obviously) defined by the LM3886 capabilities. National claim that's over 10A (11.5A, if I remember correctly). So +-10A * 4ohm = +-40Vp (80 Vpp), again much more than what the my_ref can do.
Though some people think that's an overoptimistic figure and that with +-35Vdc supply the "SpiKE" SOA protection jumps in way before that limit is reached. According to them, something like 5 - 6 amps seems to be a more realistic limit. In such a case, on 4 ohm a (really weird!) current-induced clipping would occur already at 20 - 24 Vp (40 - 48 Vpp).
And 1 Vrms ~= 2.82 Vpp x 30 = 84.6 Vpp out, much more than what the my_ref can do.
Current limit is (obviously) defined by the LM3886 capabilities. National claim that's over 10A (11.5A, if I remember correctly). So +-10A * 4ohm = +-40Vp (80 Vpp), again much more than what the my_ref can do.
Though some people think that's an overoptimistic figure and that with +-35Vdc supply the "SpiKE" SOA protection jumps in way before that limit is reached. According to them, something like 5 - 6 amps seems to be a more realistic limit. In such a case, on 4 ohm a (really weird!) current-induced clipping would occur already at 20 - 24 Vp (40 - 48 Vpp).
Blown my original MyRef A
Help needed!
I had the old version connected to neither an input or output. It was connected to power and the relais started to rattle. Afterwards it did not "amplify", the relais did not connect anymore. Probably blown the op-amp LM318 or the power-amp 3886.
Am I right?
How to analyze the fault and then solve this?
Thanks in advance
Michiel
(earsandeyes)
Help needed!
I had the old version connected to neither an input or output. It was connected to power and the relais started to rattle. Afterwards it did not "amplify", the relais did not connect anymore. Probably blown the op-amp LM318 or the power-amp 3886.
Am I right?
How to analyze the fault and then solve this?
Thanks in advance
Michiel
(earsandeyes)