It is a good idea when using a buffer circuit with a GC (or other amp) to power the buffer up first before powering the GC (or amp).
There are various ways to do this; eg using a delayed start that uses a timer and relay. Or you can use speaker protection modules that incorporate a delay before they connect the speakers.
I wondered if the following would work (and I hope you can understand the rather rough diagram
)
The idea is that when SW1 is closed power is fed to the buffer PSU. A feed is taken from the output of that supply and connected to a momentary push button switch. When that switch is closed, a relay allows mains supply voltage through to the amp PSU.
A feed taken from that PSU then keeps the relay closed after the momentary button is released.
If the mains fail, the power amps will not be reconnected until the push button is pressed.
I am not saying that this is a better solution (or even that it will work). It's just an idea that I had and wondered what the more electroncially educated think of it.
Flack jacket on.
There are various ways to do this; eg using a delayed start that uses a timer and relay. Or you can use speaker protection modules that incorporate a delay before they connect the speakers.
I wondered if the following would work (and I hope you can understand the rather rough diagram
)An externally hosted image should be here but it was not working when we last tested it.
The idea is that when SW1 is closed power is fed to the buffer PSU. A feed is taken from the output of that supply and connected to a momentary push button switch. When that switch is closed, a relay allows mains supply voltage through to the amp PSU.
A feed taken from that PSU then keeps the relay closed after the momentary button is released.
If the mains fail, the power amps will not be reconnected until the push button is pressed.
I am not saying that this is a better solution (or even that it will work). It's just an idea that I had and wondered what the more electroncially educated think of it.
Flack jacket on.
I use a relay such that the buffer output is connected via the normally closed terminals (NC) to ground when the power is off. Upon power up these contacts open up so that the relay switch is out of the circuit altogether. Capacitance is used in parallel with the relay to delay the switching of the relay after power up. This way the buffer and circuit is fully powered and stable before any signal goes to the amplifier chip. The delay required in my case is less than a second (I'm using Pedja's discreet buffer).
The current setup is a 115v dc relay, powered through a DB104 rectifier bridge, with two 47uf capacitors. You can increase the delay by increasing the capacitance. I also use two 2.2k resistors in series with the relay coil to reduce the voltage across the relay and caps below 100v (the relay coil is 15k ohms). Whole setup cost about $3.50.
Sheldon
p.s. I don't know if other buffers would cause noise when the ground is removed, but it's easy to test by just manually grounding the buffer output with the unit on.
The above approach would also solve the power off problem in most cases. It's possible that a break in power could be shorter than the time required for the relay to close (it's delayed just like the on cycle) but longer than required to discharge the main power supply caps. In that case, you might get a "thump".
The current setup is a 115v dc relay, powered through a DB104 rectifier bridge, with two 47uf capacitors. You can increase the delay by increasing the capacitance. I also use two 2.2k resistors in series with the relay coil to reduce the voltage across the relay and caps below 100v (the relay coil is 15k ohms). Whole setup cost about $3.50.
Sheldon
p.s. I don't know if other buffers would cause noise when the ground is removed, but it's easy to test by just manually grounding the buffer output with the unit on.
The above approach would also solve the power off problem in most cases. It's possible that a break in power could be shorter than the time required for the relay to close (it's delayed just like the on cycle) but longer than required to discharge the main power supply caps. In that case, you might get a "thump".
Nuuk,
Almost all of my GC amps (including Joe's Hybrid) includes a delayd relay for speaker swithing. My opamp buffered GCs needed less time while the tube buffered takes a lot longer before connecting the output. However, when the power switch is turned off, the relays are switched off within milliseconds which is another useful feature.
However, I think your circuit is a very clever one but if power is the one to be switched, the GC section (amp side) must absolutely be "thump-free" when switched on.
Cheers!
Almost all of my GC amps (including Joe's Hybrid) includes a delayd relay for speaker swithing. My opamp buffered GCs needed less time while the tube buffered takes a lot longer before connecting the output. However, when the power switch is turned off, the relays are switched off within milliseconds which is another useful feature.
However, I think your circuit is a very clever one but if power is the one to be switched, the GC section (amp side) must absolutely be "thump-free" when switched on.
Cheers!
Thanks for the feedback guys.
My aims were to avoid relays on the output of the amp and remove the problem of capacitors not discharging quickly enough in the event of a momentary power cut.
Do you mean like I have on my Velleman speaker protection modules?
Sheldon, your solutions sounds neat. Do you use just one relay, or one for each channel?
Jojo, I agree about the different length of delays needed for different buffers. The valve buffers take longest while I have found the opamp and discrete buffers need only a second or two to 'settle down'. I appreciate that my suggestion does not deal with the problem of power up thump (although I don't get that with my 'smaller' PSU).
My aims were to avoid relays on the output of the amp and remove the problem of capacitors not discharging quickly enough in the event of a momentary power cut.
Do you mean like I have on my Velleman speaker protection modules?
Sheldon, your solutions sounds neat. Do you use just one relay, or one for each channel?
Jojo, I agree about the different length of delays needed for different buffers. The valve buffers take longest while I have found the opamp and discrete buffers need only a second or two to 'settle down'. I appreciate that my suggestion does not deal with the problem of power up thump (although I don't get that with my 'smaller' PSU).
Single pole relays
If anyone's interested, i can get 24v dc coil single pole relays with contacts rated @240v / 7A...
We throw away photo cell sat work that use these, and it's never the relay that goes faulty ,only the electronics that control them !
(Cells are used to switch your street lights off and on)
If anyone's interested, i can get 24v dc coil single pole relays with contacts rated @240v / 7A...
We throw away photo cell sat work that use these, and it's never the relay that goes faulty ,only the electronics that control them !
(Cells are used to switch your street lights off and on)
To Nuuk : I don't know how it make Velleman, I have own solution . Look at http://www.sjostromaudio.com/dudek/files/PA03_00.pdf There you will see quite simply timing circuit, based on binary counter 4020 ( with quite simply protection circuit ). Timing circuit based on counter have one big advantage : if you use two ( or more ) independent counters in different amps, you will be sure, that the time will be the same. 4020 you can feed from the same supply as preamp. Time delay of this circuit can be easy changed only by option of output Q from short time ( useable for example for soft start of amp) to " middle " time ( delayed output relay ) or how you can. Problem will be not with long time for heating of tubes .
. Look at http://www.sjostromaudio.com/dudek/files/PA03_00.pdf There you will see quite simply timing circuit, based on binary counter 4020 ( with quite simply protection circuit ). Timing circuit based on counter have one big advantage : if you use two ( or more ) independent counters in different amps, you will be sure, that the time will be the same. 4020 you can feed from the same supply as preamp. Time delay of this circuit can be easy changed only by option of output Q from short time ( useable for example for soft start of amp) to " middle " time ( delayed output relay ) or how you can. Problem will be not with long time for heating of tubes .
carlosfm said:Will you guys stop fiddling with the LM3875?
Use the LM3886.
It's better IMHO, and has an integrated solution for your problem.
Makes it a bit easier, but you still have to have a switch of some sort which would give the appropriate delay. If I understand the specs. properly, the switch could be ss, so smaller than a relay and no mechanicals but, in principle, the same.
Sheldon
The link is fixed now.Nuuk said:Thanks Upopa but I just get a 'page unavailable' at present. I will try again later.
Sheldon, it took me a few hours to get familiar enough with corcuit maker to complete a whole circuit but like all software, it gets easier, the more you use it.
A hint to Pavel and everybody else: When you copy links that are above a certain length and you have "..." inside the URL you must right-click with your mouse and then choose "copy this link location" (or something like that). diyaudio chops the links when they get too long.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.