hi guys,
I was looking for a nice projects for my "opera" loudspeakers.
I have tried many solid state amps (from 50 to 150W rms) but they are too powerful for my room and lossy for my taste. No detail or scene...
Can you suggest a very very good little diy amp (20-30w max) which has awesome quality and detail?
regards
I was looking for a nice projects for my "opera" loudspeakers.
I have tried many solid state amps (from 50 to 150W rms) but they are too powerful for my room and lossy for my taste. No detail or scene...
Can you suggest a very very good little diy amp (20-30w max) which has awesome quality and detail?
regards
pass single ended class a from Audio magazine 1977 (or now downloadable from web site) sounds like what you want.
mlloyd1
mlloyd1
thank you very much...but I think original devices are not available now, and I was looking for a class AB design if possible
regards
regards
Later modern production laterals are available are available from:
Magnatec. ALFET Lateral MOSFETs
Guess that rules out my suggestion then. Also lateral FET.
I was shamelessly going to say my own Lateral FET design which has stood the test of time as regards musicality. There's good and there's very good. You mention very very good hence the recommend.
Its all here but there are no boards available. Its a true diy construction if you want to have a go.
http://www.diyaudio.com/forums/solid-state/119151-my-mosfet-amplifier-designed-music.html
too much power your design MOOLY...
I already have a 150W class AB amp that sounds really good, but can use it at no more that 5-10%...is for that reason I'm looking for something low power that I can use at all of the potentiality...something tube-like 🙂
I already have a 150W class AB amp that sounds really good, but can use it at no more that 5-10%...is for that reason I'm looking for something low power that I can use at all of the potentiality...something tube-like 🙂
VHex+
VHex+ with one pair of output devices, downsized to 30-35W output power.
Excellent amplifier, sounding great at 1W or close to clipping. It's really pretty unique, utilizing an elegant voltage gain topology. You will not experience lack of detail for sure.
Here is the build by XRK:
VHex+ in a simple compartment
See some professional measurements and feedback there, above the post down that link.
Available in the form of boards, kits, assembled/tested modules or a fully-built amplifier.
Cheers,
Valery
VHex+ with one pair of output devices, downsized to 30-35W output power.
Excellent amplifier, sounding great at 1W or close to clipping. It's really pretty unique, utilizing an elegant voltage gain topology. You will not experience lack of detail for sure.
Here is the build by XRK:
VHex+ in a simple compartment
See some professional measurements and feedback there, above the post down that link.
Available in the form of boards, kits, assembled/tested modules or a fully-built amplifier.
Cheers,
Valery
Attachments
How into diy building a design are you ? Do you need PCB's and full instruction or can you think things through and work with an untested design, because I might have just the thing ?
Simple, it uses (or originally used) just three transistors.
Adding a fourth makes it much better though.
Low power. It can use from 18 to 40 vac transformer and its single rail. That allows the possibility of dual mono construction from a dual secondary transformer.
Power output of 6 to 40 wrms depending on transformer voltage.
I can make my own pcb but can't test...don't have an oscilloscope so I would need a tested desihn...but if you like you can post your idea...we can work together
Do not understand, what you are doing now is fine; in Hi Fi listening using the amp at, say, 5% or 10% of clipping power is perfect to have great dynamic range and hardly ever reach clipping.
A smaller amp may be perfect, but a large one is none the worse just because of that, modern music reproduction has high dynamic range, not like old style tape, vinyl or FM which needed more or less compression or equalization tricks to work comfortably.
Of course, I´m the last person in the World to stop anybody from DIYing 🙂
at the cost of being repetitive I propose to work on dx trust amp.
I think it has great potentiality but needs to be worked on by someone.
It has great characteristic but something can be really upgraded....does anybody wanna try to study the schematics?
I think it has great potentiality but needs to be worked on by someone.
It has great characteristic but something can be really upgraded....does anybody wanna try to study the schematics?
Well this started out as a three transistor design using the dual lateral FET from Semelab. Unfortunately stocks of these have sold out although suppliers are saying there will be more early next year. Of course you can use a normal pair of devices... it just adds another device to the component count. Q2 and Q3 could be a single Darlington device but I found the (real not simulated) performance was so much better with a discrete darlington. The voltage reg can be the LM317 for all but the highest voltage versions and for those there is the LM317AHV which has a 60 volt rating.
The parts to the right are just a solid state relay to eliminate switch on thumps. A 'rapid' charge version using two such relays is possible whereby the initial load is a low value resistor.
The zipped folder contains a couple more images and also the simulation file and models.
Attachments
Thanks for the reminder, I recall that design with interest.
You could try Profusion for Exicon ECX10N/P20 - same thing, different brand and part number. If that's too hard also, wouldn't 2SK1058/J162 be close enough?
Thanks Ian.
Any of the ordinary single die laterals would be fine for a design like this, and even the double die types would probably make for a decent low impedance capable amp with probably only minor changes to the VAS stage necessary (if at all).
Any of the ordinary single die laterals would be fine for a design like this, and even the double die types would probably make for a decent low impedance capable amp with probably only minor changes to the VAS stage necessary (if at all).
Mooly, I think your input biasing still needs some work. Doing the math, you could end up with about 6 mV of output noise with the input open, and still 1/10 that with it shorted. Not great either way - good integrated amps are at ~100 µV. Plus, midpoint voltage tightly regulated but a presumably unregulated supply?
I'd just tap an R - D - (R||C) voltage divider between supply and ground with your 22k. Simple, low noise and provides good tracking over a fairly wide range of supply voltages. There have been other solutions but this is one of them.
I'd just tap an R - D - (R||C) voltage divider between supply and ground with your 22k. Simple, low noise and provides good tracking over a fairly wide range of supply voltages. There have been other solutions but this is one of them.
You make a good point actually and yes, the supply would be almost certainly be unregulated. I must see how that compares for ripple rejection and so on.
The complexity has grown from my first thoughts about a design like this. As they do 🙄
The complexity has grown from my first thoughts about a design like this. As they do 🙄
Member
Joined 2009
Paid Member
You can make lower power versions of more powerful amplifiers too, lower the rail voltage and possibly adjust a couple of resistors and all will be fine. But sometimes the higher power amplifier is a better place to be - because the higher voltages can mean that the BJTs are operating in a more linear fashion due to less influence from the Early Effect. And more powerful amplifiers have strong power supplies which can be helpful.
My favourite amplifier design is also Class AB and BJT (mostly, the full version has a pair of MOSFETs but they could be left out): http://www.diyaudio.com/forums/solid-state/245619-tgm8-amplifier-based-rod-elliot-p3a.html
However, this is not a beginner-friendly design to build. One day I may revisit this TGM8 and make a simpler through-hole version.
The designer of that amp has many other designs, some of which have better technical performance without an output capacitor etc., why not pick one of those.
Last edited:
- Status
- Not open for further replies.