I am considering various options for rectification in an unregulated power amp supply. There are various monolithic, high current (e.g. 35A, 50A) bridge rectifiers that one can purchase for a few dollars (I have a few of these already). I also see designs using four individual diodes, like from Murata, perhaps soft-recovery, etc. Frankly I am not sure I even am aware of all the different varieties of power diodes out there.
Can someone please provide a broad brush perspective on the pros and cons of all the different types of diodes when used for a power amp supply?
Can someone please provide a broad brush perspective on the pros and cons of all the different types of diodes when used for a power amp supply?
Probably not super helpful, but for very low voltage stuff I tend to use discrete Schottkys and for high voltage (up to 1kV) I tend to use silica carbide or UF series diodes.
I tend to use conventional bridges for 20V - 400V or so...
I have noticed over time that the discrete diodes are more reliable - I cannot remember a single instance of one failing in a properly designed circuit, while not many I cannot say the same for the bridge rectifiers I use. (Just a few failures)
I tend to use conventional bridges for 20V - 400V or so...
I have noticed over time that the discrete diodes are more reliable - I cannot remember a single instance of one failing in a properly designed circuit, while not many I cannot say the same for the bridge rectifiers I use. (Just a few failures)
Schottky diodes have a lower voltage drop, and so lower heating. Also no reverse recovery.
Many SiC Schottky diodes are available. Diodes - Rectifiers - Single | DigiKey
Last edited:
Silicon carbide allows much improved Schottky diode ratings. Here's one I ran across in a quick search.
Admittedly a bit pricey, but ok for diy. APT20SCD65K Microsemi Corporation | Discrete Semiconductor Products | DigiKey
Admittedly a bit pricey, but ok for diy. APT20SCD65K Microsemi Corporation | Discrete Semiconductor Products | DigiKey
Last edited:
I don't see any advantage with that part. It says right on the page you linked to:
Voltage - Forward (Vf) (Max) @ If 1.8V @ 20A
For a monolithic bridge like KBPC5010 at those same conditions (20A) the forward voltage drop is only 0.95V (I believe that is for the bridge, e.g. across two diodes, taken from the datasheet) and it costs 8 times less (if making a full bridge from four discrete diodes)!
Voltage - Forward (Vf) (Max) @ If 1.8V @ 20A
For a monolithic bridge like KBPC5010 at those same conditions (20A) the forward voltage drop is only 0.95V (I believe that is for the bridge, e.g. across two diodes, taken from the datasheet) and it costs 8 times less (if making a full bridge from four discrete diodes)!
Hi CharlieLaub,
Here's an unpopular opinion. The standard bridge rectifier or single diode is very highly engineered and sold in astronomical numbers. That makes them less expensive then they would otherwise be. I have not seen any fail in designs that were properly engineered. I have seen lots of failures in equipment that had reduced cooling or extremely high hot switching currents (read: stupid high value filter capacitors).
From the sounds of what you are designing, a 500W class A-B amplifier, your voltages will be above most Schottky rectifiers range. Fast recovery diodes will help to generate ringing on your supplies while you may get some of that with normal rectifiers, but this is easily controlled.
Avoid the temptation to use extremely high amounts of filter capacitance as this will cause very short conduction angles and high IR losses in everything. Using a rectifier and filter combination for each channel will help keep the required filter capacitance down to reasonable sizes. That will depend on the lowest average impedance you expect to drive. Again, be reasonable here.
Believe it or not, a little series resistance won't really hurt anything and will help reduce the hot switching current. This is good. It will reduce continuous sine wave power levels but will not actually do much to peak levels - like when you are playing music.
-Chris
Here's an unpopular opinion. The standard bridge rectifier or single diode is very highly engineered and sold in astronomical numbers. That makes them less expensive then they would otherwise be. I have not seen any fail in designs that were properly engineered. I have seen lots of failures in equipment that had reduced cooling or extremely high hot switching currents (read: stupid high value filter capacitors).
From the sounds of what you are designing, a 500W class A-B amplifier, your voltages will be above most Schottky rectifiers range. Fast recovery diodes will help to generate ringing on your supplies while you may get some of that with normal rectifiers, but this is easily controlled.
Avoid the temptation to use extremely high amounts of filter capacitance as this will cause very short conduction angles and high IR losses in everything. Using a rectifier and filter combination for each channel will help keep the required filter capacitance down to reasonable sizes. That will depend on the lowest average impedance you expect to drive. Again, be reasonable here.
Believe it or not, a little series resistance won't really hurt anything and will help reduce the hot switching current. This is good. It will reduce continuous sine wave power levels but will not actually do much to peak levels - like when you are playing music.
-Chris
There's a volt drop per element in the bridge. It's naturally much cheaper, since it's
a commodity part in one package.
This is a dual-mono 2-channel design. I'm planning to build two identical amps, each using two 62-0-62 500W transformers. I am planning to use four 22000uF 100V caps in the bank, per channel. I could do a CRC supply and bypass the bridge with small caps to reduce the effects of ringing, etc. Have 50A 100PIV bridges for the rectification, but wanted to know if there was something that would be much better for that purpose.
Hi CharlieLaub,
I would use higher voltage rated rectifiers in case you have a spike. Since increasing the voltage spec on a rectifier is fairly inexpensive, why not use 400 V rated parts?
The capacitance is reasonable for what you are doing. Using two 22,000 uF units for each supply polarity allows more flexibility in cases. The 100 VDC rating is close to exactly what you need. Have you compensated for higher line voltages? Mine read 125 VAC quite often. Yours might be low now, but if you move ...
Sounds like you are on the right track with your design. You just need to decide if other rectifiers will save you heat energy or not. Also, your drop in the bridge will have two diodes in series, so double your voltage drop values when figuring out heat dissipation or for comparisons.
Best, Chris
I would use higher voltage rated rectifiers in case you have a spike. Since increasing the voltage spec on a rectifier is fairly inexpensive, why not use 400 V rated parts?
The capacitance is reasonable for what you are doing. Using two 22,000 uF units for each supply polarity allows more flexibility in cases. The 100 VDC rating is close to exactly what you need. Have you compensated for higher line voltages? Mine read 125 VAC quite often. Yours might be low now, but if you move ...
Sounds like you are on the right track with your design. You just need to decide if other rectifiers will save you heat energy or not. Also, your drop in the bridge will have two diodes in series, so double your voltage drop values when figuring out heat dissipation or for comparisons.
Best, Chris
this is extreme, i tend to ignore. look instead at diode characteristics at anticipated operation....
500watts at a voltage of say 500 volts is just 1 ampere....
what is the forward drop at that point? and so what is the dissipation? does it conform with diode heating specs?
i made a 16 KT88 triode wired p-p amp, i used a 15 amp. 600 volt flat inline bridge, the amp is still good after more than 5 years...
imho, maximum device specs are useful to consider under short circuit conditions, then you should have enough impedance in circuit that the device will see to avoid failure...
Nothing at all, he just asked about different diode types for use in a high power amp.
I'm building a pair of amps right now, using four of these.
Vishay high speed lowloss diode module for high speed power rectification 1 pc ! | eBay
CL will want to see this. Simple, no-math transformer snubber using Quasimodo test-jig
62-0-62 is 124 volts, a 100volt piv bridge is scary to use here...
at least, i will use a 400 or or even 600 volt bridge if that were my amp...
the prices for those are still cheap comparatively...
Sorry, when I wrote the post I left out a zero on the peak inverse voltage rating! The bridge I will use is rated 1000 PIV.
The transformer is 62-0-62 secondary. When rectified that comes to around 85-90 Vdc, which is good for about 500W into 8R. There will be some sagging if I run 4 Ohm loads, but that should be OK. The amp output devices can handle the current.
The transformer is 62-0-62 secondary. When rectified that comes to around 85-90 Vdc, which is good for about 500W into 8R. There will be some sagging if I run 4 Ohm loads, but that should be OK. The amp output devices can handle the current.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.