Check the number of low power channels your AV amp. had for an idea of how much current the 25V windings can likely supply. That will tell you what you can expect in current but you should easily power a 50W/8R amplifier on the rectified 35V DC rails.
If you like something interesting that is simple but not just another brute dinosaur for driving 6" nails, have a look at Elvee's Circlophone or Lazy Cat's SSA projects. Loads of input and good reading there, with a few threads and variations.
If you like something interesting that is simple but not just another brute dinosaur for driving 6" nails, have a look at Elvee's Circlophone or Lazy Cat's SSA projects. Loads of input and good reading there, with a few threads and variations.
The need of bias for cover-up is in direct proportion to the amount of device noise, which is caused by the device capacitance discharges.
Modern low-bias
The very modern and high end Onsemi MJL0281/MJL1302 have lower device capacitance and thus lower noise, less likely to grain/naf and doesn't need much bias. Similar high end output devices include 2SC5200/2SA1943 from Toshiba and FJL4315/FJL4215, FJP5200/FJP1943, KSC5200/KSA1943 from Fairchild.
Instead of covering clumsy with bias at the expense of power supply capacity, I'd suggest to make the amplifier as efficient as possible so that your power supply can recharge as quickly and effectively as possible despite the rather limited transformer. Normally, Honey Badger monoblocs are running with 400va apiece. You're running with 250va shared between two of them. Oh my. Therefore I suggest using low noise output devices that won't need counterproductive bias settings. With that transformer, you'd probably like to avoid choking it with bias for an effect like driving with the parking brake engaged. A cap while charging is an open door to noise, so please charge them quickly.
Here's my attempt at a Honey Badger schematic revision: http://www.diyaudio.com/forums/atta...746574-diyab-amp-honey-badger-honeybadger.gif
It was specifically made to answer your question on how many and what sort of devices to use. I've attempted to constrain the bias dials so that explode isn't one of the settings, and I've tried to remove mismatch errors, mainly at the drivers.
Discussion of it is here: http://www.diyaudio.com/forums/solid-state/192431-diyab-amp-honey-badger-19.html#post3351880
Again, no need to cripple a modern amplifier with overcurrent bias. Low noise devices allow you to fine tune the amplifier to reduce reliance on bias. Not wasting device capacity will have low noise devices operating more linearly and more durably. High bias does the opposite.
P.S.
This became suddenly very interesting! Powering the heatsink as little as possible = powering the speaker as well as possible. In this case, I think that little transformer can do an excellent job.
Thanks Daniel!
Ian: The donor amp had 35Wx2 + 35Wx2 for the center and rear channels powered from the 25-0-25VAC winding.
Ian: The donor amp had 35Wx2 + 35Wx2 for the center and rear channels powered from the 25-0-25VAC winding.
More questions......it appears that the MJL0281 is non-stock from digi-key, although the most of the Fairchild devices are available. Do these require outputs matching? I'm assuming so.....
So 2 pairs of any of the above will drop right into the circuit?
So 2 pairs of any of the above will drop right into the circuit?
Mouser has them, and all of the other devices that you would need for Honey Badger and Mongrel-AX amplifiers.
Well, that should be a shoe-in for a strong 8-4R amplifier with a single TO264 pair or 2 pairs of TO247, TO3P transistors.
If you want to go the Honey Badger route, it's a lot bigger amp which really needs to be rated 150W + to make the build worth the parts, design complexity and effort worthwhile. (IMHO)
'Fine amp with a lot of care to make the build straightforward but a lot more than a 25-75W build from from the junk-box. You won't want to give that one away to the young'n. 😉
Heh heh! You would ask about PCB's...
That's difficult for the reason that you have different suppliers that I haven't even heard of, yet they might have just the product and only live a few blocks away from you.
On this Forum, you've seen a few simple designs already and there are, or have been PCB group buys for popular designs but you know how that goes - after a year or so, no more GBs and no PCBs.
For ready PCB availability and suitable documentation, I suggest the designs on this site. They cover a broad range of applications and now even 'toobs' too. There's a price list and on-line purchasing. The full and proper documentation comes with the PCBs since the schematics you see are simplified to avoid blatant rip-offs. Some will work as is, but not too well. The really popular, simple design is P3a and there are many newb threads and serious ones too here on DIYaudio, trying to dodge even the PCB cost. You can build that with MJL21193/4 and that should be plenty tough enough.
A mountain of good articles and info there too. ESP's handle here is "rode".
Have a browse and see if anything else, like also popular P101, for a nice Mosfet sound, appeals.
Try this SYMEF amplifier. Designer, Mr. Harrison is offering the PCBs at throwaway prices. I am sure this amplifier is better than most of the amplifiers from Rod Elliots site.
http://www.diyaudio.com/forums/group-buys/206358-symef-pcb-4.html
He has several other amplifiers - 1DIFFQC, HAL OPB, SYMEF etc and is very helpful in answering your questions.
http://www.diyaudio.com/forums/group-buys/206358-symef-pcb-4.html
He has several other amplifiers - 1DIFFQC, HAL OPB, SYMEF etc and is very helpful in answering your questions.
I had the same thoughts once. Then I discovered mosfets, particularly vertical types. I once thought these devices were inferior to BJT's but I have discovered how wonderful they are if used properly. So I played around with them for a while, experimenting and such, and discovered that it is possible to drive a speaker with 75Wrms sine wave using just one pair of TO-220 devices. 1 pair of TO-220 BJT's might get up to 25Wrms but they will be limited by secondary breakdown and de-rated Pd. I have discovered through my many hours of experimenting that the planer stripe fets are very tough. They are only limited in Pd by die temperature, thus if you can keep the peak die temperature down then you can suck a surprising number of Watts of power through them. Many of the planer stripe fets with equal specs are available in TO-220 and TO-247. I noticed that the only difference between those TO-220 and TO-247 is Pd and Pd de-rating, they have exactly the same die. This means increasing the heatslug of the TO-220 would essentially make the transistor bigger. Of course there is the problem that they create quite a bit of distortion when bias in class AB that must be dealt with, but here is how to do it. In this circuit I use SMD, but that is not required. I suggest following what Bob did in his paper (in link) to get the gist of it. I think you will be quite surprised by the results. I was able to duplicate Bob's results quite easily. It has completely changed my outlook on output transistors, so much in fact that I may never use BJT’s for output devices again except maybe in small amplifiers of just a few Watts.
Symef needs a lot of current to drive 2R, otherwise a wasted design on 35V or 40V rails.
OPB & 1DIFFQC use craptanium like MJ11032/3 & TIP142. If you need a motor drive - maybe.
We're all entitled to an opinion about what's better if we have indeed compared,
but let's stick with facts and the OPs requirements.
Thanks for the replies!
Ian: I've spent a fair bit of time over on Rod Elliott's site over the past few years, although I have not bought any PCBs from him. I'll take another look over there. Good education for a newbie.
Routhun: The SYMEF looks interesting, although all of the PCBs are already snapped up, and the schemo shows 200W into 2R......seems like all of the AB stuff over here are fire-breathing behemoths.......So for this project, I would also build with less outputs?
I usually hang out in the tubes forum and have built about 4-5 tube amps and a tube pre (Aikido). Last year I wandered over to the Pass Labs Forum and ended up building a mini-A, an F4 and I have a fully populated pair of PD F5 boards screaming "use me" on the bench.
This power transformer brought me over here, so I have little knowledge of the various projects floating around.
Ian: I've spent a fair bit of time over on Rod Elliott's site over the past few years, although I have not bought any PCBs from him. I'll take another look over there. Good education for a newbie.
Routhun: The SYMEF looks interesting, although all of the PCBs are already snapped up, and the schemo shows 200W into 2R......seems like all of the AB stuff over here are fire-breathing behemoths.......So for this project, I would also build with less outputs?
I usually hang out in the tubes forum and have built about 4-5 tube amps and a tube pre (Aikido). Last year I wandered over to the Pass Labs Forum and ended up building a mini-A, an F4 and I have a fully populated pair of PD F5 boards screaming "use me" on the bench.
This power transformer brought me over here, so I have little knowledge of the various projects floating around.
Hmm, its better if someone follows their own advise. Criticizing me for my "opinions" but again giving their own opinions about those amps. Its better to stick with the facts with some data to support.
Well, since my current project is a stereo module intended for home use it does not need to be hugely powerful. 75Wrms @4R X 2 is more than enough. For this amp I am using planer stripe fets from Fairchild, FPQ50N06/FQP47P06 because the Gm is fairly similar. But, the price has risen to $2 each.🙁 I found that it really doesn't matter as much as I thought it would in this circuit because of the local error amplifier gate drive circuit so I got some FQP33N10/FQP27P06, they were only $1 each.😛 "Ugh but those are switching transistors" many will say, and I will agree. But it's not the component but rather how it is used that makes a great amplifier.😉 These devices are very similar to hexfets but they tend to have a much greater dependency of Vgs vs Id upon Vds than do the cellular types. I suppose this adds an extra distortion in Gm that is dependent upon the output voltage.
However I think they tend to be more rugged.😎
However I think they tend to be more rugged.😎- Status
- Not open for further replies.