Atom Amp has been an enormous success in its first 11 months, with around 7000 amplifiers in the wild, and thousands more in production. At this rate, Atom Amp is on track to exceed 8 years worth of Objective2s in just 18 months. We are full of gratitude to have so much community support! Thank you!
We do not sit idle at JDS Labs; I despise complacency and prefer instead to constantly improve. With high success, though, small improvements are easily lost in the noise of discussion. To recap the years’ progress:
This Atom Amp update was shared at Head-Fi back in September:
Within the first months of release, customer feedback indicated a few areas for possible improvement:
1) Gate ejection mark on the volume knob [Resolved in March 2019 batch]
2) Injection molding marks on top surface [Resolving in September 2019 batch]
3) Flow marks on front of enclosure [Acknowledged]
The knob update was integrated by March. I am pleased to report that marks on the top surface have been successfully eliminated in the production batch arriving today.
There is no ETA for item (3) due to isolated customer comments and high risk of altering the mold (problem could worsen or persist, while delaying production).
Quality Control Improvements
Let’s preface with repair and return rates. Zero (0) Atom Amps have been delivered faulty in the past 30 days. Return rate is comparable to our other lines, meaning we accept single digit numbers of non-defective returns in a 90 day period for thousands of shipped amplifiers. Our production team closely monitors these statistics and is rewarded for success during performance evaluations. No one wants to see Q/C troubles–not you, not me, and not JDS Labs employees.
Alas, perfection is fleeting. We document every instance of Q/C failures reported by our customers so that we can analyze patterns and implement changes where possible. Atom Amp’s 30-day repair rate was 0.3% at release, and it took a few short months to move from 99.7% to 100% success. We’ve made the following Q/C improvements throughout 2019:
- Automated Optical Inspection: DOA amplifiers (only 3 in the first 2000) were found to be caused by insufficient inspection time, allowing an amplifier to pass before later failing due to SMT assembly errors at the AC input. Our SMT assembly team in California implemented corrective action via AOI, and we increased our final inspection time at JDS Labs.
- Fixed Stuck Buttons: Stuck buttons were addressed by altering our assembly process, and triple checking button fit during inspection.
- Improved Relay Break-In: Stuck relays accounted for most support requests. We pinpointed the reason (micro-arcing) and evaluated two solutions. Replacing a fast MLCC capacitor in the AC rectification stage with a slower tantalum version presented an easy fix, but with a risk of total failure in the event of lightning strikes or other power surges. Instead, we chose to alter our inspection process. Any Atom Amp with a stuck relay is subjected to a minimum of 1 week of break-in, followed by repeat inspection. Support requests are now at or near zero per month.
- Improved Potentiometers: A small, but vocal number of customers expressed dissatisfaction with Atom’s channel balance. I could have ignored the rare negativity and let knowledgeable community members respond helpfully as they did. After-all, Atom Amp is assembled with the same size Alps branded potentiometer as we have used for 12 years. We chose to evaluate and improve in two ways. First, our production team was trained to adopt higher standards and reject more Atom Amps based on channel imbalance. Any potentiometers bordering unacceptable are now pulled and swapped. Second, we asked Alps to set stricter tolerance gang-matching for our RK097 pots. Alps initially declined based on our aggressive goals. After further discussion, their engineers agreed to a small gang-matching improvement at higher cost.
A few visitors have inquired about a rumor that JDS Labs will release Atom Amp in a metal case. This rumor is just that.
We’ve invested considerably to our in-house machine shop, and are certain that our CNC’s cannot keep up with Atom Amp’s demand at the current price point. Even a price increase of +$50 would be insufficient to build something like Element. Yes, we could revert to an aluminum extrusion with front/rear plates like the older Objective line. But we’ve done that. It’s boring and ugly. Why raise cost to take a step backwards? How about stamping? Already been done. We decided to try injection molding for Atom for the following reasons:
- Consistency – Every unit is the same without variations in finish.
- Flexibility – More mechanical design options compared to metal.
- Scalability – We can build a million just as easy as one. Increasing production with metal demands more machines.
We’re happy with the end result because we’re able to assemble Atom Amp quite efficiently as designed, and focus expenses on genuine components and top-tier SMT assembly in the United States. That said, we’re passionate about working machined aluminum into future releases where doing so makes sense.