@alexcp Testing this build finally...I had all sorts of trouble with a new kind of enclosure I have been trying.
I gave it a listening and it sounds very clear. Great clarity. No noise floor to speak of. Decent power output, even into 600R headphones.
I noticed 2 things in this initial series of test.
1. The PSU LED behaves unusually to me. It comes on very briefly on power on and then turns off.
The output of the PSU is the expected ~ +/- 17V (more like 16.8V, but who is counting?), PGOOD is at -20V or so (from memory, don't quote me on that.)
Since the output is good, I assume all is good.
2. XFEED behaves oddly. It must be my setup somehow. I installed a switch for it to the SMD board. When off, it's off and doesn't interfere. When I flip it on, after a minute or so, the Omicron board goes into protection, the relay clicks and the load is disconnected. It doesn't come back from that state until I set the switch to off.
3. When I turn the RK27 10K pot all the way up, I noticed that it's fairly easy to make protection kick in and switch off the load. That was with a 1 kHz sine wave.
I gave it a listening and it sounds very clear. Great clarity. No noise floor to speak of. Decent power output, even into 600R headphones.
I noticed 2 things in this initial series of test.
1. The PSU LED behaves unusually to me. It comes on very briefly on power on and then turns off.
The output of the PSU is the expected ~ +/- 17V (more like 16.8V, but who is counting?), PGOOD is at -20V or so (from memory, don't quote me on that.)
Since the output is good, I assume all is good.
2. XFEED behaves oddly. It must be my setup somehow. I installed a switch for it to the SMD board. When off, it's off and doesn't interfere. When I flip it on, after a minute or so, the Omicron board goes into protection, the relay clicks and the load is disconnected. It doesn't come back from that state until I set the switch to off.
3. When I turn the RK27 10K pot all the way up, I noticed that it's fairly easy to make protection kick in and switch off the load. That was with a 1 kHz sine wave.
What do you have for the quick disconnect resistor alex mentioned we should add? It was brought up in post #201, I think I settled at ~5.6k. The easy test to determine if it's the resistor size, is to remove the PGOOD connection between the PSU and the amp board. -20V sounds like the behavior when mine was misbehaving, with 5.6k it was more like -25V.
That was it, a 6.8K fixed it.
Although with the pot wide open, it’s very easy to make it go into protection.
With XFEED enabled, it’s a little easier still. I removed the XFEED switch for now.
Mandatory shots or it didn’t happen.
Nothing went according to plan with that enclosure. First, the paint chipped at the drill sites. Then the fixing putty would not take right. Then the paint had problems. Finally, the tolerances were incorrect and it would not close as it’s designed to. 🤦♂️😭🤦♂️😭
But it’s done and doesn’t look too bad. The paint chipped again during final assembly, I’ll touch that up…or not.
Nothing went according to plan with that enclosure. First, the paint chipped at the drill sites. Then the fixing putty would not take right. Then the paint had problems. Finally, the tolerances were incorrect and it would not close as it’s designed to. 🤦♂️😭🤦♂️😭
But it’s done and doesn’t look too bad. The paint chipped again during final assembly, I’ll touch that up…or not.
The enclosure is looking great! 
In principle, it is possible to trigger protection with a low frequency (say 20 Hz) sine wave at near clipping level. I don't believe this a realistic scenario in normal use, but if it is a problem, it can be alleviated by taking the steps from post #279.
Not the intended behavior. It should come on and stay on. Need to check how the LED rectifier is wired.
This is also unusual. Looks like a cap is being charged somewhere by a small current. Need to check how that switch is wired.
This is normal. The protection is sensitive enough to detect even a slightly asymmetric clipping, and clipping is always slightly asymmetric. It is not triggered by an occasional clipping when playing music, as the protection's LPF needs time to integrate the DC component (30 milliseconds with a 1 volt step), but when it clips continuously, protection eventually flips the switch.
In principle, it is possible to trigger protection with a low frequency (say 20 Hz) sine wave at near clipping level. I don't believe this a realistic scenario in normal use, but if it is a problem, it can be alleviated by taking the steps from post #279.
This was my fault too. I typically use red for power supplies and didn’t pay attention. I swapped in a green LED and that’s fixed.
Not sure. But I removed the switch at this point, so all is well!
All good. I am not concerned by it since I listen far away from clipping and as you say real music is unlikely to trigger this problem.
Thanks, Alex, great little amp!
It sounds like a fresh glass of spring water on a hot summer day.
I love these games too! I have been playing them for 25 years now! ****, I'm getting old.
Since there has been a number of requests for more Omicron boards, I just opened a second group buy for Omicron boards and another one for balanced line receiver boards.. Let's see if there is enough unmet demand to warrant another manufacturing run. Both GBs will close at the and of the day on Saturday, September 30. If you're interested, be sure to sign up before then.
With only a couple of days before closing the second group buy for Omicron boards, here is some (updated) reference information.
First, the schematic of Omicron for the through hole board (since the SMT version add a few extras, I will post the schematic for the SMT board separately):
The schematic above implements the input impedance mod as described in the post #67.
The board also includes Omicron's DC protection and output filter:
First, the schematic of Omicron for the through hole board (since the SMT version add a few extras, I will post the schematic for the SMT board separately):
The schematic above implements the input impedance mod as described in the post #67.
The board also includes Omicron's DC protection and output filter:
The part list aka BOM for the through hole board was posted at #155. It allows fairly broad substitution, as well as opamp rolling. Even though Omicron was designed with NE5532, it works well with a broad selection of opamps. In particular, AD712, OPA1642, OPA1652, OPA1688, OPA2604, LME49860, LM4562 all have been tested to work. The opamps that are available only in SMT only can be used on the through-hole board with adapters, or directly on the SMT board.
Part list for the SMT version:
Part numbers are just examples with one exception: the Murata pulse transformer. Be careful with replacing BAV99, as a change in forward voltage drop would affect the quiescent current of the output stage.
| Qty | Value | Device | Package | Parts | Example P/N | TH/SMT |
|---|---|---|---|---|---|---|
| 1 | 78602/3MC | Pulse transformer | Murata 786 series | L1 | 78602/3MC-R | SMT |
| 2 | RS1D | Diode | SMA | D1, D51 | RS1D_R1_00001 | SMT |
| 8 | BAV99 | Dual diode | SOT23-3 | D2, D3, D4, D5, D6, D7, D9, D11 | BAV99,235 | SMT |
| 1 | BC856 | PNP BJT | SOT23-3 | Q52 | BC856B,215 | SMT |
| 2 | 2N7002 | N-Ch MOSFET | SOT23-3 | Q51, Q53 | T2N7002AK,LM | SMT |
| 2 | 33n | Polypropylene film capacitor | Radial LS=5mm | C3, C23 | MKP2C023301B00KSSD | TH |
| 4 | 1000u | Aluminum electrolytic capacitor | Radial D=10..13mm | C13, C14, C33, C34 | EEU-FC1E102E | TH |
| 8 | 47u | Aluminum electrolytic capacitor | Radial D=5mm | C11, C12, C31, C32, C9, C10, C29, C30 | EEU-FC1E470B | TH |
| 4 | 10n | Ceramic capacitor NP0 (C0G) | 1206 | C6, C26, C45, C46 | C1206C103J5GACAUTO | SMT |
| 4 | 15p | Ceramic capacitor NP0 (C0G) | 1206 | C7, C8, C27, C28 | C1206C150J5GACTU | SMT |
| 2 | 1n5 | Ceramic capacitor NP0 (C0G) | 1206 | C5, C25 | 12065A152JAT2A | SMT |
| 7 | 1u | Ceramic capacitor X7R | 1206 | C41, C42, C43, C44, C51, C52, C53 | CL31B105KBHNFNE | SMT |
| 6 | 330p | Ceramic capacitor NP0 (C0G) | 1206 | C1, C21, C4, C15, C16, C24 | C1206C331J5GAC | SMT |
| 2 | 5p6 | Ceramic capacitor NP0 (C0G) | 1206 | C2, C22 | CC1206DRNPO9BN5R6 | SMT |
| 2 | G6K-2P-Y | Relay | G6K-2P-Y | K1, K2 | G6K-2P-Y-DC24 | TH |
| 1 | 2-way | Header | 0.1 inch pitch | XFEED_SW | 22-23-2021 | TH |
| 3 | 3-way | Header | 22-11-2032 | HEADPHONES, INPUT, LINE_OUT | 22-23-2031 | TH |
| 1 | 5-way | Header | 22-11-2052 | POWER | 22-23-2051 | TH |
| 1 | GREEN | LED | 1206 | LED | 156120VS75000 | SMT |
| 1 | LM339D | Quad comparator | SOIC-4 | U51 | LM339DR | SMT |
| 2 | PZT2222A | NPN BJT | SOT223 | Q1, Q21 | PZT2222A,135 | SMT |
| 2 | PZTA2907A | PNP BJT | SOT223 | Q2, Q22 | PZT2907A,135 | SMT |
| 2 | NE5532 | Dual opamp | SOIC8 | U1, U21 | NE5532DR | SMT |
| 4 | 68 | Thin film resistor | 1206 | R13, R14, R33, R34 | RT1206FRE0768RL | SMT |
| 8 | 100 | Thin film resistor | 1206 | R20, R40, R41, R42, R45, R46, R53, R54 | RT1206FRE07100RL | SMT |
| 2 | 100k | Thin film resistor | 1206 | R56, R60 | RT1206FRD07100KL | SMT |
| 4 | 10k | Thin film resistor | 1206 | R11, R12, R31, R32 | RT1206FRD0710KL | SMT |
| 4 | 12k | Thin film resistor | 1206 | R17, R18, R37, R38 | RT1206FRE0712KL | SMT |
| 4 | 15k | Thin film resistor | 1206 | R5, R6, R25, R26 | RT1206FRE0715KL | SMT |
| 1 | 1Meg | Thin film resistor | 1206 | R55 | RT1206FRE071ML | SMT |
| 3 | 1k | Thin film resistor | 1206 | R10, R30, R59 | RT1206FRE071KL | SMT |
| 4 | 1k3 | Thin film resistor | 1206 | R1, R2, R21, R22 | RG3216P-1301-D-T5 | SMT |
| 4 | 2.2 | Thin film resistor | 1206 | R15, R16, R35, R36 | RC1206FR-072R2L | SMT |
| 2 | 20k | Thin film resistor | 1206 | R51, R52 | RT1206FRE0720KL | SMT |
| 2 | 24k | Thin film resistor | 1206 | R19, R39 | RT1206FRE0724KL | SMT |
| 4 | 2K2 | Thin film resistor | 1206 | R7,R27,R43, R58 | RT1206FRE072K2L | SMT |
| 2 | 300 | Thin film resistor | 1206 | R8, R28 | RT1206FRE07300RL | SMT |
| 1 | 33k | Thin film resistor | 1206 | R57 | RT1206FRE0733KL | SMT |
| 2 | 470k | Thin film resistor | 1206 | R9, R29 | RT1206FRE07470KL | SMT |
| 4 | 4k7 | Thin film resistor | 1206 | R3, R4, R23, R24 | RT1206FRE074K7L | SMT |
| 4 | 51k | Thin film resistor | 1206 | R47, R48, R49, R50 | RT1206FRE0751KL | SMT |
Power supply board schematic:
The only difference from the previously published schematic is R10, the bleed resistor that discharges C14 when the AC power is lost.
Power supply board outline:
The tallest components on the board are the transformer, the two large electrolytic caps, and the heatsinks. It is possible to fit the power supply board into 1U enclosure when all those components are 30mm tall or less; 25mm is also possible and gives a very comfortable fit. The updated part list will follow.
The only difference from the previously published schematic is R10, the bleed resistor that discharges C14 when the AC power is lost.
Power supply board outline:
The tallest components on the board are the transformer, the two large electrolytic caps, and the heatsinks. It is possible to fit the power supply board into 1U enclosure when all those components are 30mm tall or less; 25mm is also possible and gives a very comfortable fit. The updated part list will follow.