I can send you a few yellow and red ones to play with. Not worth a big order.
Hi, and welcome to posting in the forums.
12.7V if after a 7812 regulator is a suspicious reading. Should be closer to 12V. BC517 Darlington driver does not need a decoupling between B,C pins also. I would remove the bypass from the 517 and make sure the 7812 has at least 2.5V Vin-Vout and it regulates properly. (Z diode was a solution for an odd 5V relay system that would not trigger but for 12V relays to trigger there is much more voltage space for the 550's Z diode mode.)
Thanks for quick reply and support.
I'm a little bit frustrated. The output and 12V partition doesn't run. I unsoldered 7812, BC550, BC517 and checked them.
The input voltage is near 18V and soldering the 7812 "solo" (without BC 550 + BC 517) output voltage is 12,9 -13.0 V!!! I used different L7812 and measured same voltages.
After that I changed every part of 12V regulation (47k, 10k, A7812, BC550, BC517, 10uF Condensator, 100nF coupling, 1N4001 controlled and measured) - but it doesn't solve the problem.
The rest ist okay (+10,7V, - 10,8V, +-2 V over R1 and so on.....)
Perhaps it is better to sleep one night. If anybody has an idea - please tell me!
Many thanks
Meinolf
Many thanks Salas,
I never thought, that heat sink is so important, unless I only solded one Input-Relay.
But: after mounting a simple heat sink, the output relay starts working
. Measuring voltage over the 1N4001 beside the output relay voltage starts slowly (ca. 4 sec. later) after switching on.
Reaching 6,4V the relay sounds like "grrrrr.rrr.rr.r...." and than it's okay. Voltage climbs up to 10.56V, than finishes.
The 1N4001 beside the single Input-relay shows 13,0V ??? But it seems to be okay.
Switching off power, first of all the input relay stops, than the output.
It's a great step, and many, many thanks to Salas.
Please tell me, whether discribed "grrrr..rrrr....rrr...rr..r" is okay or not? But for the moment the saturday evening is safed
.Meinolf
Problem solved - Mez is running
...new day - new chance:
Salas you are great and know exactly what to do. The last problem has been oszillation.
Until I unsolded the BC517 I tried to use 100nF between B and C of the BC517.
Result: starting without flutter, reaching complete voltage as like as input relays.
Consequenze:
- Don't start Mezmerize without heatsink!
- Decoupling BC517 between B and C with 100nF Cap
- any question -> ask Salas
- if I can't solve a problem -> stop project and sleep one night (as my father told me 40 years ago)
Best greetings to the forum and I will send news after combining Mezmerize and Symasym (or F5).
Meinolf
...new day - new chance:
Salas you are great and know exactly what to do. The last problem has been oszillation.
Until I unsolded the BC517 I tried to use 100nF between B and C of the BC517.
Result: starting without flutter, reaching complete voltage as like as input relays.
Consequenze:
- Don't start Mezmerize without heatsink!
- Decoupling BC517 between B and C with 100nF Cap
- any question -> ask Salas
- if I can't solve a problem -> stop project and sleep one night
(as my father told me 40 years ago)Best greetings to the forum and I will send news after combining Mezmerize and Symasym (or F5).
Meinolf
There is 0.37 Henry with about 1kOhm internal resistance in the 12V Omron or similar parts so it would not be impossible if a bit different Ft or something between brands and batches to chance more prone to oscillate, even momentarily, but it has not been reported before with 517 that is why it seemed weird to me. That extra B,C pins capacitor cut on useless bandwidth severely so it put things back in order of course. Those I got have BC517 K916 face print and measure HFE 32000 at 5mA Ib and 1.4V VBE.
Thank you, that's a very generous offer. I've already placed my order and hopefully I can sort things out. If my new LEDs don't work out, I may have to pester you.
Salas, I want to build this amp from Juma and switch it in the circuit only when I need the extra gain. I will build them on a veroboard and like to have some capacitors for decoupling of the + and - 10 V on the veroboard. Can I use 220uF, 100uF or 10uF Silmics for this, or am I loading the output too much? With the additional caps ?
There is on the Mez PCB right now a 100uF Silmic over the MOSFETs outputs.
FWIW, I received some yellow and green LEDs in a recent Mouser shipment. Mixing and matching the reds, yellow and greens, I get the following (matched into two groups for each side of the board (2xtriplet, 2xquintet)): LED Triplet: yellow (1x1.952) yellow (1x1.947) red (1x1.733) total = 5.632V LED Quintet: green (2x2.032) red (3x1.732) total = 9.26V So, given my measurements not being at the exact ideal temp in the datasheet, error in the DMM, etc., this is as close as I can get to the target of 5.64V for the triplet and 9.4V for the quintet. To get any closer I'd have to buy more LEDs. Good enough to call it a day?
Another LED question. Looks like the power LED on the board has a 2.2K current limiting resistor. I'd like to use an orange LED for the power indicator but, with a 9V battery and a 2.2K resistor, it's blindingly bright. An 8.2K resistor in place of the 2.2K tames it a bit. Is the 2.2K resistor slot flexible? If so, how high could I go?
Post 1362 (for myself when I have to go back to find this again): http://www.diyaudio.com/forums/pass-labs/176723-mezmerize-dcb1-building-thread-137.html However, I'm not sure I understand what that trick is and what it accomplishes exactly. I have a matched quad of 2SK170BL, around 10 mA, the other 6 unmatched are in that range as well (actually 3 matched pairs). Am I supposed to do something to these JFETs? Just when I think I understand and am ready to build, I get more confused.
To cut a long story short, normal way in Mez subjectively helps low current builds by introducing some rail asymmetry. Relaxing matching goals also. Total symmetry like standard in the hot-rod Hypno takes the trick. Around 200mA both ways sound the same. But it won't interfere with your LEDS symmetry you worked for anymore. Just do the thing with the middle leg soldered to its side leg in the photo. On that specific FET only.