It's the help and advice we all receive with such generosity, that makes us want to give a little something back.
Here is a shot of my own Fetzy enclosure slowly taking form. It's not quite done yet, and yet it easily took 10x-20x the time soldering up the Fetzy took. The big hole in the middle is for a pair of "VU" meters. There's a thread over in Construction Tips in case you're interested or want more pics.
And yes Hugh(i hear yours, and others, words about proper ventilation as i write this), there are some issues with this enclosure, ventilating the main enclosure space is one for sure.
The original plan called for some cutouts above and below the amp boards similar to the front panel cutout with perforated aluminium, but i was unable to source this aluminium cheaply in time, so for now i'm going to make the lid hinged, for showing off the insides, and also keeping it slightly open to allow for ventilation while i mull the aluminium situation.
Here is a shot of my own Fetzy enclosure slowly taking form. It's not quite done yet, and yet it easily took 10x-20x the time soldering up the Fetzy took. The big hole in the middle is for a pair of "VU" meters. There's a thread over in Construction Tips in case you're interested or want more pics.
And yes Hugh(i hear yours, and others, words about proper ventilation as i write this), there are some issues with this enclosure, ventilating the main enclosure space is one for sure.
The original plan called for some cutouts above and below the amp boards similar to the front panel cutout with perforated aluminium, but i was unable to source this aluminium cheaply in time, so for now i'm going to make the lid hinged, for showing off the insides, and also keeping it slightly open to allow for ventilation while i mull the aluminium situation.
Attachments
Last edited:
I offer my apology to Andrew; since my broken brain I tend to mistake the meaning of many posts........ harrumph, should keep away!! But I'm addicted......
Kris, your amp would be find if you put elegant ventilated slots above each heatsink. OTOH, you like at a high latitude, not like Oz, were maximum temperatures in summer rise to 45C. My environment is a little different to yours!
Hugh
Kris, your amp would be find if you put elegant ventilated slots above each heatsink. OTOH, you like at a high latitude, not like Oz, were maximum temperatures in summer rise to 45C. My environment is a little different to yours!
Hugh
Andrew, thanks for your input. Indeed that is a clever and very elegant way. The chimney effect should ensure that there is quite a lot of airflow through the main enclosure with that solution. And giving a very clean look from the front.
Hugh, yeah ambient temp-wise i'm in a much better spot than the more equatorial latitudes you find yourself in, but the machine will sit in a somewhat enclosed space, and will be turned on most of the time, and i do want great reliability out of it. If i can lower the temp the caps sit at by just a few degrees, without taking a major hit in appearance, it's worth it for me.
While i have both of you on the line like this, what would you reckon the effect of a 0.47uF input coupling cap would be? Perceivably lesser low bass in your estimation? I have some Jensen copper-in-oil of that value left over from a tube amp project that never materialized that i'm dying to push into service.
Most humbly yours,
Kris
Hugh, yeah ambient temp-wise i'm in a much better spot than the more equatorial latitudes you find yourself in, but the machine will sit in a somewhat enclosed space, and will be turned on most of the time, and i do want great reliability out of it. If i can lower the temp the caps sit at by just a few degrees, without taking a major hit in appearance, it's worth it for me.
While i have both of you on the line like this, what would you reckon the effect of a 0.47uF input coupling cap would be? Perceivably lesser low bass in your estimation? I have some Jensen copper-in-oil of that value left over from a tube amp project that never materialized that i'm dying to push into service.
Most humbly yours,
Kris
No problem Kris,
I have never found much subjective difference in bass performance with caps from 0.47uF to about 10uF feeding at least a 22K impedance. It certainly seems to show up in the simulations, but not too obvious at all in the listening position.
If you buy a cap like a Jensen, you HAVE to use it...........
Duktig,
Hugh
I have never found much subjective difference in bass performance with caps from 0.47uF to about 10uF feeding at least a 22K impedance. It certainly seems to show up in the simulations, but not too obvious at all in the listening position.
If you buy a cap like a Jensen, you HAVE to use it...........
Duktig,
Hugh
I found F-3dB @ ~ 10Hz quite noticeable in the effect that turn over had on the bass output, with both small and large speakers.
@ 5Hz the effect was quite subtle, almost unnoticeable.
By the time I was down below 2Hz any further changes could not be heard by me.
As a result I adopted 1.7Hz (~90ms) for the F-3dB for the input filter I put in all my wideband amplifiers.
In these tests I have to ensure the amplifier is capable of passing the LF accurately. This requires the NFB roll off and any other filters "inside" the amplifier to be set even lower, so that the input filter changes are what is being heard or not heard.
F-3dB . . . RC time constant
20 . . . . . . 8 milli-seconds (ms)
10 . . . . . . 16 ms
5 . . . . . . . 32 ms
2 . . . . . . . 80 ms
1.6 . . . . . 100 ms
1.0 . . . . . 159 ms
To obtain your desired F-3dB select the RC that matches your frequency.
The two input component values are multiplied to arrive at your RC
1uF and 100k are 100ms
0.47uF and 100k are 47ms
2u2F and 20k are 44ms
0.47uF and 22k are 10.3ms
In another Thread I discovered that I have been ignoring the effect of Source resistance (Rs) in the above F-3dB calculation.
Fortunately, I always use a low Rs and can ignore the error.
the actual formula for determining the roll off frequency of the input filter is:
F-3dB = 1 / { 2 Pi (R+Rs) C}
If Rs <<< R then one can ignore it.
If Rs is greater than R/5 then it must be taken into account.
@ 5Hz the effect was quite subtle, almost unnoticeable.
By the time I was down below 2Hz any further changes could not be heard by me.
As a result I adopted 1.7Hz (~90ms) for the F-3dB for the input filter I put in all my wideband amplifiers.
In these tests I have to ensure the amplifier is capable of passing the LF accurately. This requires the NFB roll off and any other filters "inside" the amplifier to be set even lower, so that the input filter changes are what is being heard or not heard.
F-3dB . . . RC time constant
20 . . . . . . 8 milli-seconds (ms)
10 . . . . . . 16 ms
5 . . . . . . . 32 ms
2 . . . . . . . 80 ms
1.6 . . . . . 100 ms
1.0 . . . . . 159 ms
To obtain your desired F-3dB select the RC that matches your frequency.
The two input component values are multiplied to arrive at your RC
1uF and 100k are 100ms
0.47uF and 100k are 47ms
2u2F and 20k are 44ms
0.47uF and 22k are 10.3ms
In another Thread I discovered that I have been ignoring the effect of Source resistance (Rs) in the above F-3dB calculation.
Fortunately, I always use a low Rs and can ignore the error.
the actual formula for determining the roll off frequency of the input filter is:
F-3dB = 1 / { 2 Pi (R+Rs) C}
If Rs <<< R then one can ignore it.
If Rs is greater than R/5 then it must be taken into account.
There are 3 adjustment pots.
They need to be adjusted in the correct order.
Then you need to go through and correct them again, because there is some interaction between the stages as one pot is adjusted.
You may have to repeat 3 times to finalise your settings.
Be careful you don't overheat any device due to excessive current. This can come about because the pots were set to the wrong position before start up, or because you turned the pot in the wrong rotational direction.
You need to clip on your measurement probes, so that you can see what effect each adjustment is having.
They need to be adjusted in the correct order.
Then you need to go through and correct them again, because there is some interaction between the stages as one pot is adjusted.
You may have to repeat 3 times to finalise your settings.
Be careful you don't overheat any device due to excessive current. This can come about because the pots were set to the wrong position before start up, or because you turned the pot in the wrong rotational direction.
You need to clip on your measurement probes, so that you can see what effect each adjustment is having.
They should be at least low inductance types, as are the specified current sense resistors . Many good quality, low value (<1R) ceramic block wire wound resistors are actually bifilar wound and this reduces inductance considerably below plain wound types.
Wirewound resistor » Resistor Guide
http://www.arcolresistors.com/wp-content/uploads/2012/07/ACS-7-12.30.pdf
Wirewound resistor » Resistor Guide
http://www.arcolresistors.com/wp-content/uploads/2012/07/ACS-7-12.30.pdf
Thanks Andrew... I have to order some missing parts firstly .... especially SMD power resistors is little bit difficult to order in my countryQuestion, have to be 5W 0R15 resistor mettalic or can be wirewound ?
These SMD power resistors are a part of the Power supply filtering. Within the space constraints, any type of resistor will do (I think)
You are right gannagi, but knowing Hugh's PCB designs, he would strive for the best audiophile quality boards possible at a reasonable cost. These flat resistors are standard parts in his own products and the higher component specifications are probably the main reason the project remains popular.
Thank you Ian, a positive comment!
A few pointers:
I have tried to use 1% 600mW Beyschlag through hole resistors. Anything from Vishay (who own most of the passive component industry) are fine. Beyschlags are now obsolete, but of course metal film resistors live forever in storage.
The decoupling 0R15 resistors are Vishay Dale 4512 WSR2 smds. These are rated to 2W continuous, but they are expensive and difficult to find. I use them because the layout mandates very thin, large dissipation devices UNDER the pcb. I have not found anything else to the the job.
I specify 3,300uF caps on the power supplies. Marcel, I notice you are using 6,800uF KMH from Nippon Chemicon, and these will work well BUT put a higher surge rating on the rectifiers, so I suggest a BYV32E ultrafast soft recovery double diode package, since the handle higher surge than the BYQ28E recommended.
You are doing a good job, and you will be a great amp which is very suitable for efficient speakers and works well with entertainment systems.
Cheers,
Hugh
A few pointers:
I have tried to use 1% 600mW Beyschlag through hole resistors. Anything from Vishay (who own most of the passive component industry) are fine. Beyschlags are now obsolete, but of course metal film resistors live forever in storage.
The decoupling 0R15 resistors are Vishay Dale 4512 WSR2 smds. These are rated to 2W continuous, but they are expensive and difficult to find. I use them because the layout mandates very thin, large dissipation devices UNDER the pcb. I have not found anything else to the the job.
I specify 3,300uF caps on the power supplies. Marcel, I notice you are using 6,800uF KMH from Nippon Chemicon, and these will work well BUT put a higher surge rating on the rectifiers, so I suggest a BYV32E ultrafast soft recovery double diode package, since the handle higher surge than the BYQ28E recommended.
You are doing a good job, and you will be a great amp which is very suitable for efficient speakers and works well with entertainment systems.
Cheers,
Hugh
Thank you Hugh,
I like filtration capacity little bit oversized(mainly I bought caps for super price) .. I used BYV42-200, I hope, should by fine ... as I can see from other pictures, heatsink is mount on just one rectifier.. also C5,C6 was not placed (i don´t know what difficult load mean)
I like filtration capacity little bit oversized(mainly I bought caps for super price) .. I used BYV42-200, I hope, should by fine ... as I can see from other pictures, heatsink is mount on just one rectifier.. also C5,C6 was not placed (i don´t know what difficult load mean)
Attachments
Last edited:
Today I connect amp to my testing PSU . I already set bias ,offset, and 24V across R12.. But i have some problem..sine look fine, but square is, how can I say...Look at the pictures... I have no more time, so tommorow I will continue....
Attachments
What frequency, Marcel?
However, the FetZilla is not particularly good at higher frequency square wave, BUT, it should be good at 1KHz.
Rather than worrying too much about square wave, which is not exactly music, how does it sound now? Remember, it was design specifically for a good subjective presentation, with no concern for tight square waves, which have harmonics eleven times the fundamental and VERY difficult to pass through an amplifier.
Cheers,
Hugh
However, the FetZilla is not particularly good at higher frequency square wave, BUT, it should be good at 1KHz.
Rather than worrying too much about square wave, which is not exactly music, how does it sound now? Remember, it was design specifically for a good subjective presentation, with no concern for tight square waves, which have harmonics eleven times the fundamental and VERY difficult to pass through an amplifier.
Cheers,
Hugh
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.