most builders who value their speakers would fuse the power lines.
This is what limits the time and value of DC current that can flow into your speaker when something goes seriously wrong.
For 27W into 8ohms try using F2A or F1.6A
for 50W into 4ohms try using F2.5A or F2A
I am not sure whether I want one rail to be still on while the fuse of the other is blown, for whatever reason.
I agree that for safety reasons there should be a mechanism to switch off the rails. But I rather have both rails switched off at the same time.
And switching off the rails is still too little to save your speakers, as you still need to drain off all the stored energy in the PSU caps first.
Just my 2 cents,
Patrick
I agree that for safety reasons there should be a mechanism to switch off the rails. But I rather have both rails switched off at the same time.
And switching off the rails is still too little to save your speakers, as you still need to drain off all the stored energy in the PSU caps first.
Just my 2 cents,
Patrick
Fused rails are also not my piece of cake.
A DC-protection board with a fast differential amp (with fast I mean response time of amp and low pass filter) and a high quality Omron relay works excellently for me.
Can be simply cooked together from D. Selfs DC-sensor amp and Nelsons slow on/quick off circuit (a la XOno). I think I posted this circuit here sometime.
Has the nice advantage of a delayed turn-on as well
A DC-protection board with a fast differential amp (with fast I mean response time of amp and low pass filter) and a high quality Omron relay works excellently for me.
Can be simply cooked together from D. Selfs DC-sensor amp and Nelsons slow on/quick off circuit (a la XOno). I think I posted this circuit here sometime.
Has the nice advantage of a delayed turn-on as well
The fuse is after the main smoothing caps.
Yes, I agree a mechanism for detecting one rail "off" and then blow the other fairly quickly would be better.
If rail voltage is ~25Vdc and speaker resistance is ~6r0 then the 4Adc will blow the F1.6A fuse in ~~10 seconds or so. A smaller fuse would be safer but we need to avoid nuisance blowing.
How about using a crowbar circuit that smokes everything to save the speaker?
Actually the foldback circuit should save the outputs if it's working properly.
Snubbers and fusing the power lines
Hi,
well it looks like I will not be fusing the power lines. I am using a circuit from ESP in Australia (their P33). This is a relay cutoff driven by a frequency dependent filter. I need this as the F5 in one mode will be a full range amp and in another mode will be the driver of the mids / highs. As it will be driving either an NS1000 or a JBL4343 it is also frequency switchable for the actual crossover frequency of the specific speaker. So, based on the input so far, no fuses in the power lines.
Did you notice the resounding silence to my other question concerning the use of snubber caps on the rectifiers? I am going to interpret this as meaning "optional" as there (unusually for this forum) appear to be no strong feelings on this ... or am I wrong?
George
Hi,
well it looks like I will not be fusing the power lines. I am using a circuit from ESP in Australia (their P33). This is a relay cutoff driven by a frequency dependent filter. I need this as the F5 in one mode will be a full range amp and in another mode will be the driver of the mids / highs. As it will be driving either an NS1000 or a JBL4343 it is also frequency switchable for the actual crossover frequency of the specific speaker. So, based on the input so far, no fuses in the power lines.
Did you notice the resounding silence to my other question concerning the use of snubber caps on the rectifiers? I am going to interpret this as meaning "optional" as there (unusually for this forum) appear to be no strong feelings on this ... or am I wrong?
George
Many years when I built my first amp I was told to mount very smal caps on each diode on supply bridge, 4 in total, and values were small nano or pico farad
snubbers I understand as slightly bigger bypass caps combined with a resistor, to alter Q, or slow it down if you like
back then I used film bypass caps too
until I tried to remove them again
a bit mellow without
but clearly "ringing" with the bypass caps
never used them since
btw, in another thread "best electrolytis", theres a discussion on the subject how bypass cap interacts with supply electrytic caps
not sure. but it may have been Leach who also wrote something about bypass film caps
he said the bypass cap size and type should be found by measurements only, and never chosen randomly
As 'Tinitus' mentioned above, snubbers are R-C networks - there is a good, if old, intro on Jim Haggermans site about these.
The snubber network (R-C plus C) directly after the bridge is for reducing the diode 'noise' only - you still need filters to reduce the ever increasing "line noise" and physical layout to avoid the diode current pulses radiating from the transformer, wiring, etc from getting into the amp itself.
The typical "block bridge" is rather unsatisfactory for this, and you can now get good Shottky bridges that are much better.
You can also use a technique called "tuning the transformer" that gives remarkable results (you do need extremely remarkable patience!) and info about this is on "Promitheus" website.
There is another approach to the diode "problem" introduced by John Brown (Ultimate NOS 1541A dac ... website) using three diodes in series with 2 resistors to reduce the diode current pulse to "merely annoying" size.
And then there is the choice of wire between transformer and diodes (multistrand), diodes to first cap,(included small resistor) and the choice of that first cap, regardless if using a choke or resistor for the C-R-C, C-L-C, etc method
There's quite a bit of effort in getting the power supply right in any amp, and with the F5, substantial improvements can be obtained with a little thought, some $s, some extra power, heat, and a lot of perserverance.
... my 2 cents.
The snubber network (R-C plus C) directly after the bridge is for reducing the diode 'noise' only - you still need filters to reduce the ever increasing "line noise" and physical layout to avoid the diode current pulses radiating from the transformer, wiring, etc from getting into the amp itself.
The typical "block bridge" is rather unsatisfactory for this, and you can now get good Shottky bridges that are much better.
You can also use a technique called "tuning the transformer" that gives remarkable results (you do need extremely remarkable patience!) and info about this is on "Promitheus" website.
There is another approach to the diode "problem" introduced by John Brown (Ultimate NOS 1541A dac ... website) using three diodes in series with 2 resistors to reduce the diode current pulse to "merely annoying" size.
And then there is the choice of wire between transformer and diodes (multistrand), diodes to first cap,(included small resistor) and the choice of that first cap, regardless if using a choke or resistor for the C-R-C, C-L-C, etc method
There's quite a bit of effort in getting the power supply right in any amp, and with the F5, substantial improvements can be obtained with a little thought, some $s, some extra power, heat, and a lot of perserverance.
... my 2 cents.
So, a crowbar with say 10 amp fuses in the rail?
What is the crowbar?
Solid state device?
Relay?
Will it actually work in practice?
Does anyone really care, or need it?
These and other questions, "inquiring minds want to know"!
_-_-bear
PS. and what do Crows do with bars? Do you need a heavy metal bar to move a crow? Is it a place where large black birds go to imbibe alcoholic beverages, or pick up other crows?? Listen to crow music??
What is the crowbar?
Solid state device?
Relay?
Will it actually work in practice?
Does anyone really care, or need it?
These and other questions, "inquiring minds want to know"!
_-_-bear
PS. and what do Crows do with bars? Do you need a heavy metal bar to move a crow? Is it a place where large black birds go to imbibe alcoholic beverages, or pick up other crows?? Listen to crow music??
Trimming pots P1, P2
Slightly off the present topic, but ....
As the pots (P1,2) are in parallel with the "main gain" resistors (R3,4 on original article Figure 6) will the "sound" of the basic cermet pots affect the sound of high quality resistors at this crucial point in the signal path?
No doubt, this has been covered before, and can anyone indicate approx post in this long thread?
Also, as the Thermistors decrease impedance with temperature, the "colouration" of these devices in parallel with the main resistors and pots assume more significance, yes?
The apparent improvement of the sound by leaving out the thermistors makes one wonder if the pot can be also "improved" on?
Perhaps those "pesky little LDRs" from trhe Lightspeed thread can be pressed into service here for trimming and temp compensation?
Slightly off the present topic, but ....
As the pots (P1,2) are in parallel with the "main gain" resistors (R3,4 on original article Figure 6) will the "sound" of the basic cermet pots affect the sound of high quality resistors at this crucial point in the signal path?
No doubt, this has been covered before, and can anyone indicate approx post in this long thread?
Also, as the Thermistors decrease impedance with temperature, the "colouration" of these devices in parallel with the main resistors and pots assume more significance, yes?
The apparent improvement of the sound by leaving out the thermistors makes one wonder if the pot can be also "improved" on?
Perhaps those "pesky little LDRs" from trhe Lightspeed thread can be pressed into service here for trimming and temp compensation?