I don't know, I asked about CRC filter equations - I got indirect answers. I asked about NTCs - I was told to read a book (check the thread). Is that what this forum is for? I have a library across the street, I thought the internet was interactive.
To put this issue in perspective, I've had numerous PMs from folks saying they had the same issues as I had with the "Balanced" thread. That's amazing to me. People in the forum, afraid to post their questions because they feared ridicule, frustrated due to run-around answers, tired of "inside" communications in the thread...
I guess I thought DIYAudio was about sharing experiences, not a place to attempt to demonstrate superiority. I'm early into this hobby, but if I learned something substantial, either by experience or luck, I'd publish every aspect of it to anyone who'd listen. Yet some folks learn something new and met it out, bit by bit, to squeeze every last drop of being "cool" out of it. That's not my cup of tea. Life's too short, I'll build a Cascoded F5, and when the Balanced Amp get's fully published (6 months from now...maybe) I'll drag my parts out of the closet and finish it.
None the less, I'll test the "Cascoded F5" on the weekend.
To put this issue in perspective, I've had numerous PMs from folks saying they had the same issues as I had with the "Balanced" thread. That's amazing to me. People in the forum, afraid to post their questions because they feared ridicule, frustrated due to run-around answers, tired of "inside" communications in the thread...
I guess I thought DIYAudio was about sharing experiences, not a place to attempt to demonstrate superiority. I'm early into this hobby, but if I learned something substantial, either by experience or luck, I'd publish every aspect of it to anyone who'd listen. Yet some folks learn something new and met it out, bit by bit, to squeeze every last drop of being "cool" out of it. That's not my cup of tea. Life's too short, I'll build a Cascoded F5, and when the Balanced Amp get's fully published (6 months from now...maybe) I'll drag my parts out of the closet and finish it.
None the less, I'll test the "Cascoded F5" on the weekend.
Last edited:
when are you going to let go of this annoying continious complaining
time to look forward, and not back
maybe you should have built "standard" F5, as described in manual, instead of doing something you are not yet able to understand
F5 is very carefully documented, and almost 100% safe to build
sure, we all do stuff we may not fully understand
but then there is no other than one self to blame
lets meet here in a couple of years, and share experiences
btw, I have been reading and studying F5 thread for a very long time, and first now I feel ready to build
if that means anything
I supose I am one of the peoples that sent the PM
Basicaly I agree with NYCOne
I also agree with Tinitus
I would not like to wait but I could not experiment (wich I am willing to do) and blow my speakers
So here we go
Hi Zen
I have one if you could...
I am thinking about a mosfet used as switch to cut off the supply to the Amplifier PCB after the capacitor bank as a way to implement speaker’s protection.
Question is supposing 28 V 4 A what would be the demand in terms of heath dissipation?
The mosfet would be I believe in the saturation region showing a very small resistance
I was also thinking of using LM339 to detect the signal crossing the "zero" and a set reset timer.
If the DC offset is to hi or one of the mosfets get short circuit the timer does not reset and either disconnect the speakers or switch off the sur mentioned mosfet.
I am stuck with the timing circuit could you help please?
Tanks
Al
Basicaly I agree with NYCOne
I also agree with Tinitus
I would not like to wait but I could not experiment (wich I am willing to do) and blow my speakers
So here we go
Hi Zen
I have one if you could...
I am thinking about a mosfet used as switch to cut off the supply to the Amplifier PCB after the capacitor bank as a way to implement speaker’s protection.
Question is supposing 28 V 4 A what would be the demand in terms of heath dissipation?
The mosfet would be I believe in the saturation region showing a very small resistance
I was also thinking of using LM339 to detect the signal crossing the "zero" and a set reset timer.
If the DC offset is to hi or one of the mosfets get short circuit the timer does not reset and either disconnect the speakers or switch off the sur mentioned mosfet.
I am stuck with the timing circuit could you help please?
Tanks
Al
I think that someone more xperienced with timing circ will chime in ;
in any way - I decided long ago that best speaker protection is crowbar on output ( to gnd)
like in Quad 405
everything above that , in complexity , is above my level of interest ...... if we exclide simple ones with relays in line with speaker
I don't mind having relay in line with speaker .......
in any way - I decided long ago that best speaker protection is crowbar on output ( to gnd)
like in Quad 405
everything above that , in complexity , is above my level of interest ...... if we exclide simple ones with relays in line with speaker
I don't mind having relay in line with speaker .......
I've built a few standard F5s, if you read the F5 thread, you'd have seen my posts.
You also seem confused, I have not started a single complaint, save the first post. All my other responses were answering other's posts.
Finally, I thought the idea if this forum was to try new things and help each other learn. Do you think it's something else?
You also seem confused, I have not started a single complaint, save the first post. All my other responses were answering other's posts.
Finally, I thought the idea if this forum was to try new things and help each other learn. Do you think it's something else?
when are you going to let go of this annoying continious complaining
time to look forward, and not back
maybe you should have built "standard" F5, as described in manual, instead of doing something you are not yet able to understand
F5 is very carefully documented, and almost 100% safe to build
sure, we all do stuff we may not fully understand
but then there is no other than one self to blame
lets meet here in a couple of years, and share experiences
btw, I have been reading and studying F5 thread for a very long time, and first now I feel ready to build
if that means anything
ZM,
Thanks for offering help (again). As long as you asked about my questions with the Balanced F5...
I was unsure of the my power supply. I was trying to understand what it would take on the CRC filter and NTC inrush to make the PS work like the stock F5. I was considering two 500 VA (4x15V each) transformers from Antek. The thread had suggested a 0R68 for the CRC.
My questions are:
Thanks for offering help (again). As long as you asked about my questions with the Balanced F5...
I was unsure of the my power supply. I was trying to understand what it would take on the CRC filter and NTC inrush to make the PS work like the stock F5. I was considering two 500 VA (4x15V each) transformers from Antek. The thread had suggested a 0R68 for the CRC.
My questions are:
- Can one use the resistor in the CRC filter to make the output voltage land where you want it? The thread suggested 16V rails. Can I "get" the rails to 16V by variying my R until the desired voltage is reached?
- The Antek 5415 has four outputs, I only need two. Do I just use two and leave two unconnected, or do I tie two in parallel (twice, to form two outputs from the four)?
- With inrush current, NP ultimately suggested CL-60 or CL-30s. I was worried about the steady state current and the resistor value of the NTC. I wanted to know if Ametherm's SL22 10008 was ideal (Ametherm SL22 10008 ---10 ohm / 8 Amp Inrush Current Limiter Data Sheet). It's 10R0 and has a steady state current rating of 8A. It can handle 90J, it it too robust?
- Finally, how big a C does one need in the CRC filter? I saw 20000uf for each C suggested at one point, but it seemed low to me.
clarify #2 - four wires(two independent secondaries) or four independent secondaries ( 8 wires ) ?
'll be back later , maybe not before night ...
I am not ZM. But I thought I would give you my take on your questions.
First read http://firstwatt.com/pdf/art_power_supplies.pdf
Second, Download PSUD2.
Yes. There may be better ways if the difference is more than a few volts.
Either will work. I would parallel 2 for lower IR losses in the transformer.
Bigger has more turn on surge. I would look at PSUD2 and balance ripple, turn on surge, Q (look at what happens when the power supply changes current demands.)
HTH
Doug
Its not a good idea to use resistors in the supply to drop voltage because they will either catch fire, or at least will be burning a lot of heat. It may be fine in preamps, but not in a class A power amp where you have a large constant current draw. If 16 V rails are suggested and you are getting 18V, I don't think it should matter much. NP has said many times that all voltages in schematics are approximate.
The power dissipation is 'I' squared 'R'. Measure the voltage drop across the resistor, divide by the ohms of the resistor and that gives you the current through the resistor. Square the current and multiply by R and this gives you the power that must be dissipated in the resistor.
If you're passing 2 amps through a 0.5 ohm resistor, the dissipation is 2x2x0.5 = 2 watts. I would recommend atleast a 10 watt resistor for this duty as it is continuous dissipation.
A 'C' of 20,000 uF is freakin huge... obviously you can have bigger caps, but make no mistake, this is a big cap capable of killing you if you're not careful.
If it were me, I'd build the stock F5 first. This is too ambitious for a first project (again, for me).
so .....
1. as other members advise you already - don't try to burn more than ,say,500mV at R part in CRC ( or multiple) filter ; reasons - excessive dissipation on resistors , same as unnecessary slowing down entire PSU .
download PSUD from Duncan's , read help - where is explained thoroughly how to determine all things , especially how to include necessary data for xformer - internal impedance etc.
2. use one diode bridge per each secondary , then parallel two and two ;
point is in paralleling after the bridge ; look for principle in Papa's FW PSU ; reasons - no buzz and no current hogging if secondaries aren't identical in smallest amount ; and they're identical only if they're wound multifillar ..... which isn't smartest way for making cheap tororids
ask for help/exact schm if you're unsure in this
I have feeling that you're good with 15Vac secondaries - but your homework is to bring us numbers ;
you can always post *.psud file here ( zip it ) , and I'll (or someone else ) gladly double check for you
3. 10R thermistor is good for 220/240V mains , maybe overkill for you - as you're in States ; use CL60 - it's easy available there , at least to my knowledge
4. I'm using 33mF-0R1-33mF per polarity for Papa's Koan amp ;
more precisely - 800VA Donut , 4x18Vac ; so - per channel : two 18Vac - two bridges - 4 33mF caps , two 0R1 resistors
easiest path for every Greedy Boy - look at Papa's original schm - augment VA & C for your liking , put brain on neutral , enjoy .
Last edited:
While I agree with ra7 here, there are other ways of achieving this.
First you can parrallel four 3W x 2 Ohm resistors.
So that will be 0.5^2 x 2 =0.5W per resistor.
The other option is CRCRC or CRCRCRC etc
Eg 15000uF 15000uF .25Ohm 15000uF 15000uF .25Ohm 15000uF 15000uF
Then you can parrallel four 3W x 1 Ohm resistors.
So now you have 0.5^2 x 1 = 0.25W per resistor.
This is just to show how you could safely go about it. However I wouldn't worry about dropping the voltage if you are in the ball park. This amp will go into class B so there is no worries about dropping the bias a bit to compensate for increased voltage.
Last edited:
Yes, this is possible too. As long as you can provide enough margin to cover the dissipation needs, you should be fine.
The point is to not drop volts on the resistor, coz the dissipation starts getting out of hand pretty quickly. I know because I've smoked some resistors myself. If you need to drop a lot of volts, for example, if you're at 25 V and you need 16 V, it is best to use a regulator like in ZV3.
The point is to not drop volts on the resistor, coz the dissipation starts getting out of hand pretty quickly. I know because I've smoked some resistors myself. If you need to drop a lot of volts, for example, if you're at 25 V and you need 16 V, it is best to use a regulator like in ZV3.
I don't think the actual equations are likely to advance the
discussion much, but here's a graphic from an F5 supply
showing the DC and AC voltage seen by the amplifier
channel with three values of resistance in the CRC (the
C's being 30,000 Uf). The upper trace shows no
resistance, the middle trace 0.25 ohms (stock value) and
the bottom trace 0.5 ohms.
As resistance is added, the noise gets lower and there
is greater reduction of high order harmonics in the noise,
but you start to lose rail voltage.
Attachments
I may be wrong, but if you wanted 16V rail after rectification, wouldn't you want a 12V output from the transformer? such as Antek - AN-5412
If I remember correctly, the calculation for Dc voltage after rectification is 1.4. So at 15V from the trafo, you would get 21Vdc after rectification and CRC. At 12V from the trafo you would get 16.8V.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.