Now everything is perfect. If you look at all the additional soldering, you'd inevitably conclude that the board isn't really that good.
Attachments
Last edited:
Now I just need to see if the amp lasts longer than a year. 🧐
I'd like to make one more mod: the VAS transistors! These are the 4 MPSA that control the drivers. The TO-92 case has an idle temperature of over 40°C. I think that's because the preamplifier is running in Class A mode. I'd like to get this temperature lower.
I'd like to make one more mod: the VAS transistors! These are the 4 MPSA that control the drivers. The TO-92 case has an idle temperature of over 40°C. I think that's because the preamplifier is running in Class A mode. I'd like to get this temperature lower.
Yes, thanks for the tip. I swapped the gate resistance values. I always thought the 2sj type had the higher resistance. I've seen that before with other MOSFET amps... But apparently that's not the case.
Yes, that's right, you already reported on this. I mixed something up and looked at the bias trimpot. But the error is in the DC offset. Sorry.
Last edited:
No problem, It took me quite some time to find the the DC trimpot issue on the PCB. How is your amplifier performing with the new resistor values?
To be honest, it's just as good as before. However, I'm still listening in mono with a high-quality full-range speaker. There might even be big, noticeable differences if I were to test it in stereo with my high-quality floorstanding speakers. A balanced sound alone isn't enough for me. I'm striving for a wider, holographic soundstage. So a new test phase is now beginning for me. I want to listen in stereo and also swap the VAS transistors from MPSA to FZT. The FZT were used in the Goldmund Telos. The MPSA are currently too hot for me anyway. Mimesis and Telos are essentially the same. However, Telos has lower current in the VAS stage. This should make the sound more detailed. The input stage, however, remains Mimesis because it provides a wider soundstage than Telos. The current is higher there than Telos. That was my motivation for changing the resistors to how they were set in the original Mimesis.
By the way, I was wondering what inspired Goldmund to name their amplifiers. I found out that Mimesis, which comes from the Greek, means "in imitation of..." So if we're building a clone, then the name is apt. But what is Goldmund trying to emulate? Telos, which translates as "end," like Goldmund's Next Gen (Next Generation) series. They're further developments or improvements. To me, that means specifically that they've expanded their hot drivers by adding more drivers to reduce the load... They've also changed the currents in the preamps because it makes a difference in the sound. Which is ultimately a matter of taste. Improving the detail of the previous model certainly has its justification, I think. But as always, some get changes better and other get worse. The idea is for the DIY enthusiast to clone a hybrid and enjoy a $100,000 amplifier.
Please don't take this text too seriously, as it is my interpretation of things. This information does not come from Goldmund or anyone else.
Please don't take this text too seriously, as it is my interpretation of things. This information does not come from Goldmund or anyone else.
Last edited:
I am aiming for low distortion and nice soundstage, but reliabilty is first on my agenda for now. I now have 4 pcb's and am waiting for more components before I can continue the build. I plane to change 2 boards to look more like the Sphinx Myth 11 just to see if the different driver stage would make a difference. The Sphinx circuit is very similar and has different compensation in the VAS. These are good boards to experiment with. I not sure of Golmund's naming but they have a good reputation so to seem to a goid project for diy.
Your goal is good. I've looked at the Myht 11 circuit and wonder who cloned it from whom. 😉 I'm looking forward to your tests and feedback on the changes. As for the circuit board, I'm almost inclined to etch my own. What's holding me back is the lack of time. I'm too creative and full of audio projects. I'll be busy with them for the next 10 years. That's why I'm sticking with the Hm3s. They're really versatile, inexpensive, and the amp doesn't sound any worse for all the tweaking on the circuit board. No, quite the opposite.
What surprises me is that the VBE has a TO-92 transistor. It may be correctly dimensioned both electrically and thermally. I always thought the bias transistor was also a temperature compensator and belonged on the heatsink of the power MOSFET. With a TO-92, it's going to be a bit of a tinkering job. Furthermore, the Toshiba lateral MOSFETs
require one temperature compensator more than the Hitachi lateral ones. But it's certainly a very good amplifier design. Even though I didn't hear.
require one temperature compensator more than the Hitachi lateral ones. But it's certainly a very good amplifier design. Even though I didn't hear.
Attachments
Last edited:
My Hafler DH 200 does not have the bias resistor on the heatsink. It will depend on the mosfets type.
Yes, right. I knew that much too. Hitachi, Exicon and Renasas don't need a temperature compensator. With IRFP end Toshiba Mosfet things are different.
Our Hm3s board has one. uselessly? Nevertheless, I used it by mounting the VBE transistor on the heatsink of the MOSFET transistors. I'll give it a try and unscrew the VBE transistor and see what the Bias current changes after it warms up. Does it change little or nothing? Then we can solder the Transistor onto the board and that's it. Otherwise, mounting it on the large heatsink is a small improvement.
Our Hm3s board has one. uselessly? Nevertheless, I used it by mounting the VBE transistor on the heatsink of the MOSFET transistors. I'll give it a try and unscrew the VBE transistor and see what the Bias current changes after it warms up. Does it change little or nothing? Then we can solder the Transistor onto the board and that's it. Otherwise, mounting it on the large heatsink is a small improvement.
Last edited:
I really need to do a necropsy on some of my boards here, but I had assumed it was the drivers failing since they tend to run hot, nearly 50c in my setup. Assumptions mean nothing however until proven…
Running at 43v alone doesn’t seem like it should be testing anything.
Perfect amp other than the reliability.
I have a set of matching Profusion outputs for a future project so am hopeful!