Hello, I have 2 sets of A75 boards and 8 pairs of Hafler 500 heatsinks. I also have 2 Dynaco ST400 power trans formers. The transformers came from 200watt stereo amplifiers and have 55 volt secondaries. I do not know the current ratings of the transformers. Can these transformers be used with the Pass A75. If so what changes need to be made? My plan is to build 2 amps and bridge them for higher output. Thanks, Don Nebel
Don,
I did something similar, but not quite so extreme. I used a 42V secondary transformer that gives me 58VDC loaded. Your transformer will give you around 77VDC unloaded, good enough for 300+W if you increase the front end voltage to around 85V to allow 70V output swing.
That means you'll be running at nearly twice the design voltage. You'll want to run through the whole design and check currents and dissipation.
For example, R11 will now see 76V, and far exceed 1/4 W. Increase it to 20K or so and verify that you have enough current in Z1. The input differential transistors will see 85V, so they'll dissipate nearly .4W. You can bend leads and use IRF610/9610 there, which would be a good idea at dissipation.
Q7 and Q8 will need healthy heat sinks - dissipating ~.7W. Q9 and Q10 will be dissipating 1.7W. There really isn't room on the board for enough heatsink.
If you bias as specified, you'll end up dissipating 350 w/channel and still only get 75W class A into 8 ohms. Your DH500 heat sinks will handle it, if you keep the fan speed up. If you reduce the bias to 100 mA per device you are probably safe.
Lower than that you need to worry about thermal compensation. I moved Q11 on top of one of my output devices (I used IRFP240/9240) to provide compensation. It seems a little overcompensated, since the bias now takes several minutes to come up when cold. I suspect mounting Q11 on the heat sink would be better, but I haven't been motivated enough to tear them down again.
The front end regulators need the same treatment - changing the feedback resistors will be required to get 85V. You'll also need at least 100V caps. I would reccomend using a separate 40V transformer to supply the front end rather than try to find 200V caps and enough heat sink to keep the regs alive if you double the 55VAC.
All told, I think a Leach amp would be easier to modify to run on 77V than an A75. Change a couple resistors, the MJE15030's to MJE15032s and use MJ21195/6 outputs.
BTW, I'm guessing that you want to bridge your amps for sub duty. An A75 is wasted on sub only use. Save the heat, use a lightly biased class AB amp for the sub and A75s at around standard voltage for the mains.
I did something similar, but not quite so extreme. I used a 42V secondary transformer that gives me 58VDC loaded. Your transformer will give you around 77VDC unloaded, good enough for 300+W if you increase the front end voltage to around 85V to allow 70V output swing.
That means you'll be running at nearly twice the design voltage. You'll want to run through the whole design and check currents and dissipation.
For example, R11 will now see 76V, and far exceed 1/4 W. Increase it to 20K or so and verify that you have enough current in Z1. The input differential transistors will see 85V, so they'll dissipate nearly .4W. You can bend leads and use IRF610/9610 there, which would be a good idea at dissipation.
Q7 and Q8 will need healthy heat sinks - dissipating ~.7W. Q9 and Q10 will be dissipating 1.7W. There really isn't room on the board for enough heatsink.
If you bias as specified, you'll end up dissipating 350 w/channel and still only get 75W class A into 8 ohms. Your DH500 heat sinks will handle it, if you keep the fan speed up. If you reduce the bias to 100 mA per device you are probably safe.
Lower than that you need to worry about thermal compensation. I moved Q11 on top of one of my output devices (I used IRFP240/9240) to provide compensation. It seems a little overcompensated, since the bias now takes several minutes to come up when cold. I suspect mounting Q11 on the heat sink would be better, but I haven't been motivated enough to tear them down again.
The front end regulators need the same treatment - changing the feedback resistors will be required to get 85V. You'll also need at least 100V caps. I would reccomend using a separate 40V transformer to supply the front end rather than try to find 200V caps and enough heat sink to keep the regs alive if you double the 55VAC.
All told, I think a Leach amp would be easier to modify to run on 77V than an A75. Change a couple resistors, the MJE15030's to MJE15032s and use MJ21195/6 outputs.
BTW, I'm guessing that you want to bridge your amps for sub duty. An A75 is wasted on sub only use. Save the heat, use a lightly biased class AB amp for the sub and A75s at around standard voltage for the mains.
Thank you for the response. Sounds like It would be more practical to stick to the original design as far as transformers go. My interest in more power is because I have Acoustat Spectra 1100 electrostatic speakers. I also have a pair of Audio Concepts Sub 1 subwoofers. I am currently using 2 Dynaco 400 amplifiers in a Bi-amp setup to power the speaker system. I use 1 amplifier to drive the electrostatic panels and anoter amplifier to drive the woofers and subwoofers. I had read in a previous post by Mr. Pass that 12 pairs of output devices are overkill in the A75. How many devices (IRFP240/9240) would be required for a 2 ohm load?
Without running the Safe Operating Area numbers I'd stick with at least 12 pairs for your load. Pass said 6 pairs of the metal IRF2340/9240 devices was enough, but given that your load is low impedance and reactive, it is harder on the output stage. You'll have high voltage and high currents at the same time.
You might consider reducing the voltages to allow higher bias (more 2 ohm class A power) with less heat. I wouldn't go below 25V output rails, though based on Pass' comments about the IRF devices needing that much. This will give you 120W RMS into 2 ohms. You'll need over 7 amps bias to get 100W in class A. This is 354W dissipation @ 25V rails - at 12 pairs you're at 15W per device. You'll need a fair amount of airflow in your tunnels to make that work reliably. Class A operation into low impedance loads is hard to do.
This amp does quite nicely in heavily biased class AB. (despite NP's comment that he "pities the fool who tries to make a class AB MOSFET amp.") I originally had mine biased at 200 W per channel. I use a pair to biamp my speakers, plus a DH-500 for my subs. The heat was too much in the summer. I reduced the bias to 120W/channel and it still makes my living room 5-10 degrees warmer than other areas after a couple hours. Keep in mind Pass' direction that you should use at least 100 mA per output pair bias for thermal stability.
You might consider reducing the voltages to allow higher bias (more 2 ohm class A power) with less heat. I wouldn't go below 25V output rails, though based on Pass' comments about the IRF devices needing that much. This will give you 120W RMS into 2 ohms. You'll need over 7 amps bias to get 100W in class A. This is 354W dissipation @ 25V rails - at 12 pairs you're at 15W per device. You'll need a fair amount of airflow in your tunnels to make that work reliably. Class A operation into low impedance loads is hard to do.
This amp does quite nicely in heavily biased class AB. (despite NP's comment that he "pities the fool who tries to make a class AB MOSFET amp.") I originally had mine biased at 200 W per channel. I use a pair to biamp my speakers, plus a DH-500 for my subs. The heat was too much in the summer. I reduced the bias to 120W/channel and it still makes my living room 5-10 degrees warmer than other areas after a couple hours. Keep in mind Pass' direction that you should use at least 100 mA per output pair bias for thermal stability.
use of Dnyaco power xformers
Don, when I first built my A-75 about 10 years ago, I also used the exact same transformers, they worked ok but ran very hot and hummed louder than a helicopter. I tried everything like tightening the screws and laminanting etc. Finally I gave up and bought a torrodial two sizes bigger than the one spec't in the originial article and no more hum. I run dual mono up to the transformer and repalced the diode bridges with the new fet types. The amp has a lot more grunt now and is very quite. Once in a while a get a bit of turn on noise in one channel but it is probalbly a connection somewhere. My amp is 12 years old and is still my primary amp. I also built the Aleph 1.2's and this fall I will attempt an X but a high power version. Good luck. dave
Don, when I first built my A-75 about 10 years ago, I also used the exact same transformers, they worked ok but ran very hot and hummed louder than a helicopter. I tried everything like tightening the screws and laminanting etc. Finally I gave up and bought a torrodial two sizes bigger than the one spec't in the originial article and no more hum. I run dual mono up to the transformer and repalced the diode bridges with the new fet types. The amp has a lot more grunt now and is very quite. Once in a while a get a bit of turn on noise in one channel but it is probalbly a connection somewhere. My amp is 12 years old and is still my primary amp. I also built the Aleph 1.2's and this fall I will attempt an X but a high power version. Good luck. dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.