It is a matter of how much ripple current that you are allowing on your voltage rails. The higher the capacitance on the rails, the smaller the 60Hz voltage that will appear on them. The Leach amplifier has a high Power Supply Rejection ratio meaning that the circuitry can compensate a varying power supply voltage without the rail voltage signal appearing on the output as a signal. However high, there is still a finite amount which will appear on the output. So I always choose the higher capacitance at the risk of a higher volume occupance taken by the caps. My stereo Leach SuperAmp has one 55Kmfd per voltage rail. The stereo pair shares rails. Also, the higher the capacitance, the lower the rail droop will be under massive loading conditions.
BeanZ
BeanZ
You are abolutely right!!! The lights do dim and occasionally if you catch it on an AC line zero-crossing, the breaker will trip. This is only because there is inadequate inrush current limiting. I use a very small switch for power (actually two switches). There is a smaller transformer inside my amplifier which is turned on first and which provides power for accessory circuitry such as LED VU meters, fans and fan controllers, and clipping indicators, and more importantly coil power for a small relay which is then turned on by the second small power switch. When this relay is energized, the contacts trigger the gate on a TRIAC which runs in the 1st and 3rd quadrants. The TRIAC is the main power switch which can be controlled with more attractive smaller, lower current switches preventing the use of a huge industrial rated power switch. In-rush current limiting will be added sometime in the future. I am a switching power supply designer and will eventually replace this large unregulated supply.
For less experienced players, in-rush current limiting is a MUST with this size of power supply.
For less experienced players, in-rush current limiting is a MUST with this size of power supply.
My Leach amplifier is a dual-mono chassis, with two 30,000 uF capacitors per rail for a total of a quarter-Farad at nearly 60 volts.
Ripple voltage is gratifyingly low, and obviated any need for a regulated supply.
Needless to say, the room lights dimmed noticably when I turned
on the amplifier, and I was concerned about burning up the power switch contacts and blowing the rectifiers.
My simple fix was to use a double-pole center-off switch; one side has a 6 ohm 10 watt power resistor in series, the other side is
straight through. The Start position limits inrush current, and
the Run position of course doesn't; using both poles in parallel reduces stress on the contacts.
There are thermistors useful for this purpose; I simply used what I had in my collection. A 25 watt aluminum cased resistor bolted to the chassis would not be overkill, especially if you left the switch in 'start' position and proceeded to party all night at maximum volume.
Ripple voltage is gratifyingly low, and obviated any need for a regulated supply.
Needless to say, the room lights dimmed noticably when I turned
on the amplifier, and I was concerned about burning up the power switch contacts and blowing the rectifiers.
My simple fix was to use a double-pole center-off switch; one side has a 6 ohm 10 watt power resistor in series, the other side is
straight through. The Start position limits inrush current, and
the Run position of course doesn't; using both poles in parallel reduces stress on the contacts.
There are thermistors useful for this purpose; I simply used what I had in my collection. A 25 watt aluminum cased resistor bolted to the chassis would not be overkill, especially if you left the switch in 'start' position and proceeded to party all night at maximum volume.
Hey BeanZ.
I used to be a power supply jockey for Lucent Technologies. I'm just a tech though.... The 2 projects I worked on were for Telecomm. and were 208VAC in 24V 100A out and 208VAC in 48V 50A out. We had PFC AC-DC supplies on the input and DC-DC on the output..... Let us know if you get a well regulated and protected supply from 110VAC input going!!!! We used full bridge topology! I don't know about the PFC section as I was responsible for the DC-DC side..........
Good luck!
Chris
I used to be a power supply jockey for Lucent Technologies. I'm just a tech though.... The 2 projects I worked on were for Telecomm. and were 208VAC in 24V 100A out and 208VAC in 48V 50A out. We had PFC AC-DC supplies on the input and DC-DC on the output..... Let us know if you get a well regulated and protected supply from 110VAC input going!!!! We used full bridge topology! I don't know about the PFC section as I was responsible for the DC-DC side..........
Good luck!
Chris
Diode:
In days long gone, I worked for Solid State Systems in Georgia;
we had lots of fun designing switchmode power supplies for
our telephone systems. I did all the test and repairs on the
power supplies, but alas, I didn't learn much about the
technologies. While there was a dramatic reduction in size
and weight over the ferroresonant-based linear supplies, getting
the switchmode supplies reliable took a couple of years.
Resonant mode converters looked interesting, though. But
business took a turn for the worse, the Class D ringer amp
got shelved, and I finally was laid off.
That ended my career in electroncs.
Certainy there are commerical amplifiers out there with switchmode power supplies, but experimenting with those
ideas has a necessarily low priority with me. I'm still very interested in the concepts.
In days long gone, I worked for Solid State Systems in Georgia;
we had lots of fun designing switchmode power supplies for
our telephone systems. I did all the test and repairs on the
power supplies, but alas, I didn't learn much about the
technologies. While there was a dramatic reduction in size
and weight over the ferroresonant-based linear supplies, getting
the switchmode supplies reliable took a couple of years.
Resonant mode converters looked interesting, though. But
business took a turn for the worse, the Class D ringer amp
got shelved, and I finally was laid off.
That ended my career in electroncs.
Certainy there are commerical amplifiers out there with switchmode power supplies, but experimenting with those
ideas has a necessarily low priority with me. I'm still very interested in the concepts.
I have already built a four channel class-T integrated amplifier which does 190WRMS / Channel into 4 ohms. The power supply on this was a custom switching power supply which utilized an active Power Factor Correction circuit and a Zero-Voltage switching phase-shifted full-bridge converter stage which ran directly from the 400VDC bus generated by the PFC preregulator.
The rails are around 45V.
I have mentioned this project on this site many, many times I WILL form a website as I have promised so many people showing the fundamental construction and design of switching amplifiers and power supplies. I will also discuss design of the switching power supplies and regulators..step by step by step. If you are interested in reading more about this you can do a search on this site for: Class-T and BeanZ. Feel free to email me with any questions concerning this. Again, I will be making details available on a website sometime in April....and it will be announced on this site.
Here, the triac is used as a first and third quadrant switch. A simple RC snubber across the switch limits the switching noise to a very acceptable level. A triac is a pair of anti-parallel SCRs with a common gate.
The rails are around 45V.
I have mentioned this project on this site many, many times I WILL form a website as I have promised so many people showing the fundamental construction and design of switching amplifiers and power supplies. I will also discuss design of the switching power supplies and regulators..step by step by step. If you are interested in reading more about this you can do a search on this site for: Class-T and BeanZ. Feel free to email me with any questions concerning this. Again, I will be making details available on a website sometime in April....and it will be announced on this site.
Here, the triac is used as a first and third quadrant switch. A simple RC snubber across the switch limits the switching noise to a very acceptable level. A triac is a pair of anti-parallel SCRs with a common gate.
I'd sure like this information too!!!!!!
Please Please Please.... Don't put up a cool project that we simple people can't get parts for. I hate magazine articles that boast some big, cool project but the ONLY company that makes say the core of a switching transformer has a 500.00 minimum and the list goes on...... If it isn't feasable Please don't get us all hyped up, only to be let down.
This is all in respect but a lot of dissappointment behind this post. I hope you understand my concept here.
Maybe you could make it scaleable to different voltages that different projects require!
Don't get me wrong, I'd LOVE this project.
Chris
Please Please Please.... Don't put up a cool project that we simple people can't get parts for. I hate magazine articles that boast some big, cool project but the ONLY company that makes say the core of a switching transformer has a 500.00 minimum and the list goes on...... If it isn't feasable Please don't get us all hyped up, only to be let down.
This is all in respect but a lot of dissappointment behind this post. I hope you understand my concept here.
Maybe you could make it scaleable to different voltages that different projects require!
Don't get me wrong, I'd LOVE this project.
Chris
The Tripath modules originally used for this project, TA0102A, have been replaced with the TA0104A. This part is supposed to be able to emulate any of the previously discontinued parts making them a pin for pin replacement. This would be one of the most difficult parts to get. However, getting in touch with Tripath for parts never gave me a problem and I was always able to get small quantities. The core materials for some of the switching magnetics may be difficult too but I plan on providing vendor contacts and information for this as well. All of the other parts are readily available through Digi-Key, Allied Electronics, and/or Mouser. There should be no problems for US folks but for over-seas folks, I am not familiar with your part distributor counterparts but I am sure that they are not "exotic" parts requests. Stay tuned at the end of April / Early May for a website showing how this type of project is achieved.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.