Wow!
And you're using the plain industrial BOM, am I right?
Consider at least Amtrans in place of FKP2s, they're worth every cent too
Hi,
i'm going to integrate my MyRef FE with an arduino based LDR volume control, another project developed here.
When I turn on the system, there's a couple of seconds where Arduino boots and the system isn't in a stable state: I can clearly hear a buzz noise for a second or two then a bump and after that the noiuse is over and everything seems to work properly.
To avoid this, I would like to know if it's possible (and if yes, how) to delay the switching of the relay that turns on the MyRef modules.
I remember that "soft start" subject was previously discussed here. I tried to find the related posts but i haven't found anything useful yet. Do I have to read more carefully?
Any further suggestion is welcome, thank you
Best regards,
Filippo
i'm going to integrate my MyRef FE with an arduino based LDR volume control, another project developed here.
When I turn on the system, there's a couple of seconds where Arduino boots and the system isn't in a stable state: I can clearly hear a buzz noise for a second or two then a bump and after that the noiuse is over and everything seems to work properly.
To avoid this, I would like to know if it's possible (and if yes, how) to delay the switching of the relay that turns on the MyRef modules.
I remember that "soft start" subject was previously discussed here. I tried to find the related posts but i haven't found anything useful yet. Do I have to read more carefully?
Any further suggestion is welcome, thank you
Best regards,
Filippo
Filippo,
I would be careful with changing the relay time constant.
Don't forget that it has (at least) two tasks: antibump on turn on but also disconnect in case of DC on output. If you change the -unique- time constant, it will turn off too slow, I'm afraid..
It would be a better idea to insert, sum to the RC timing an external 'disable' signal, that you control with the the same Arduino?
Ciao, George
I would be careful with changing the relay time constant.
Don't forget that it has (at least) two tasks: antibump on turn on but also disconnect in case of DC on output. If you change the -unique- time constant, it will turn off too slow, I'm afraid..
It would be a better idea to insert, sum to the RC timing an external 'disable' signal, that you control with the the same Arduino?
Ciao, George
I agree that's an easy answer. I've been known to use the timer between my ears as a delay. ;-)
Years ago, I did some playing with the speaker protection circuit. It has been too long to remember details, but my memory tells me that the larger capacitor, 100 uF is involved with the turn on delay and the smaller 22 uF cap is for protection. I agree with Joseph K, there may be interactions so you need to understand it before changing anything. This would be a great circuit to simulate and play with values. Even the free online simulations should work for this.
Jac
opamps in FE
Guys,
It's quite a time that I had started to play with different opamp substitutions in our FE and Myref Evolution amps. It's like one year already on..
I had started principally with fet input opamp families, becasue as stated had found a nice going easy match with the myref topology. When I left off I've been just about to test some bjt input opamps.
That time I have finished testing the LME49990. I can confirm that it was Not a nice, easy going match in the Myref topology.. Had to sweat quite a lot.
When I finished fighting with compensation, I was rewarded with results like this: first to left is with original LM318, right is with LME49990
Guys,
It's quite a time that I had started to play with different opamp substitutions in our FE and Myref Evolution amps. It's like one year already on..
I had started principally with fet input opamp families, becasue as stated had found a nice going easy match with the myref topology. When I left off I've been just about to test some bjt input opamps.
That time I have finished testing the LME49990. I can confirm that it was Not a nice, easy going match in the Myref topology.. Had to sweat quite a lot.
When I finished fighting with compensation, I was rewarded with results like this: first to left is with original LM318, right is with LME49990
Attachments
The sound of a wire
As it could be expected, the distortion reached impressive low levels (in the high frequency range) with the new opamp.
But but. I was so dis-impressed by the amount of fight needed to achieve this, that I did not even proceed to complete two boards and give it a listen..
I beg your pardon..
In the meantime, the *627 family amps had been giving such a satisfaction in our setups that I have been asked quickly to convert all four of our Evolution amps to the configuration ADA4627, or, in my case, the OPA827.
Practically this is the configuration that in our small group we are using these amps in tha past ~~ one year.
We had made comparative listening at each time, at each swap out of the LM318. All time we concluded to proceed.. And no going back since.
I'm not saying that the original would be bad. We just found to get more satisfaction with the substitutes.
And some days ago I have just found a new element going in favour of this configuration, with the *627 family.
This time I had put the original OPA627 in the FE test modules. These are the ones which I liked so much with the ADA4627. Well, the OPA627 is no slouch as well. Going by memory, maybe it's less of a 'magic' than the ADA4627 had been, a bit more 'factual', dry, less spectacular, but very musical with fantastic resolution of details.
But but.. again, as it already happened before, I had experienced a little small step back from the usual sound, like a bit of a 'collapse' of everything, a (slight, mind you) sense of compression. With respect to my reference, the full modded, dual current pump Evolution.
And as before, I don't feel like the known differences would justify this.. So I was searching, again, to find out tha cause which I had been chasing, in vain, already before.
Then I arrived to the feedback cap, the Black Gate PK which I have in there.
It's there because I had found it the best tradeoff in the past (thanks again, Dario!), Cerafines included..
The question is : why do we need a cap in that location?
Short answer is: because of the input bias currents of the LM318, plus the topology of the amplifier. Which results not compensated for the offeset generated by the drop in the different input impedances seen by the input pins.
The original myref puts out ~800mV output offset due this dis-balance in the input impedances, if no feedback cap is applied, or a servo circuit.
But what is the case with the new family of fet opamps? They exhibit many orders less input bias current with respect to the bjt input LM318. As a side effect, also the input current noise is orders smaller, and in fact one can observe a decrease in the noise levels with these amps..
The facts: a short circuit in place of the feedback cap C9 in my FE modul with OPA627 resulted in 2mV / 3mV output offset, Left/Right channel.
As a matter of fact, with fet input opamps one does not need anymore to compensate for offset.
And the sound is: Wov. I think I have found the missing link for that old problem lingering from the past.. That feedback cap, any kind of it, is a great holdback for good sound, unfortunately..
In fact, I have just completed the 'full cirle' : had taken off, disconnected also the DC servo loop in the Evolution -- with the same very nice result in the sound opening up.. (and below 1mV DC for both channels this time).
All the best,
George
As it could be expected, the distortion reached impressive low levels (in the high frequency range) with the new opamp.
But but. I was so dis-impressed by the amount of fight needed to achieve this, that I did not even proceed to complete two boards and give it a listen..
I beg your pardon..
In the meantime, the *627 family amps had been giving such a satisfaction in our setups that I have been asked quickly to convert all four of our Evolution amps to the configuration ADA4627, or, in my case, the OPA827.
Practically this is the configuration that in our small group we are using these amps in tha past ~~ one year.
We had made comparative listening at each time, at each swap out of the LM318. All time we concluded to proceed.. And no going back since.
I'm not saying that the original would be bad. We just found to get more satisfaction with the substitutes.
And some days ago I have just found a new element going in favour of this configuration, with the *627 family.
This time I had put the original OPA627 in the FE test modules. These are the ones which I liked so much with the ADA4627. Well, the OPA627 is no slouch as well. Going by memory, maybe it's less of a 'magic' than the ADA4627 had been, a bit more 'factual', dry, less spectacular, but very musical with fantastic resolution of details.
But but.. again, as it already happened before, I had experienced a little small step back from the usual sound, like a bit of a 'collapse' of everything, a (slight, mind you) sense of compression. With respect to my reference, the full modded, dual current pump Evolution.
And as before, I don't feel like the known differences would justify this.. So I was searching, again, to find out tha cause which I had been chasing, in vain, already before.
Then I arrived to the feedback cap, the Black Gate PK which I have in there.
It's there because I had found it the best tradeoff in the past (thanks again, Dario!), Cerafines included..
The question is : why do we need a cap in that location?
Short answer is: because of the input bias currents of the LM318, plus the topology of the amplifier. Which results not compensated for the offeset generated by the drop in the different input impedances seen by the input pins.
The original myref puts out ~800mV output offset due this dis-balance in the input impedances, if no feedback cap is applied, or a servo circuit.
But what is the case with the new family of fet opamps? They exhibit many orders less input bias current with respect to the bjt input LM318. As a side effect, also the input current noise is orders smaller, and in fact one can observe a decrease in the noise levels with these amps..
The facts: a short circuit in place of the feedback cap C9 in my FE modul with OPA627 resulted in 2mV / 3mV output offset, Left/Right channel.
As a matter of fact, with fet input opamps one does not need anymore to compensate for offset.
And the sound is: Wov. I think I have found the missing link for that old problem lingering from the past.. That feedback cap, any kind of it, is a great holdback for good sound, unfortunately..
In fact, I have just completed the 'full cirle' : had taken off, disconnected also the DC servo loop in the Evolution -- with the same very nice result in the sound opening up.. (and below 1mV DC for both channels this time).
All the best,
George
Last edited: