Paul Buff
Offline
• • • •
Upload & Sell: Off
|
I'll go through this one more time, at the cost of getting slammed for "dodging blame" and being called arrogant, eccentric and a foolish business person.
I am accustomed to having 100% control over my designs and products, and have a good record of projecting my work outcome in terms of time frames and performance. They were all (except for RR1) and Compuscene) all analog, and they for the most part set the stage in 1986 for the modern monolight offered by competitors.
I designed Einstein, Cyber Commander and CyberSync in 2007. After hiring the now-infamous Chad Pryson to assist my Chief Engineer, Mike Morgan with the digital programming aspects, which are not in my are of expertise, I soon found Chad was nowhere near up to the task. So I turned to a consulting company - one very good at all things digital, to do the programming, some of the rf and circuit board layout, etc. In the process, the consulting firm thought they had ideas on how to employ switching power supplies as a means of achieving better efficiency and the avoidance on the need for pure sine inverters. I had long earlier successfuly protyped and tested the IGBT flashtube and the current power supplies, and done all the mechanical designs and molded housing designs . . . molds and metal parts have been made for well over a year.
I worked with them, and with a power supply engineer on their staff for about a year. There developed a good degree of professional jealousy during this time between myself and the power supply designer over what would work and what wouldn't. I deferred to their expertise on the issue and let them do their thing. But In the end, I was correct and they were never able to deliver a power supply that came close to the requirements.
At that point, I was under attack from forums for delays in what should have taken 4-6 months to complete, so I designed a scaled down "AB Max", employing the portions of the switching power supply that they claimed were solved and workable. Time proved that the scaled power supply design was not even up to the standards I demanded. (I actually have a near-production-ready ABMax, but I concluded it was, again, not nearly close enough for me to put my name on it and ship it.) So I declared an end to the idea of using a switching supply at all, and designed and prototyped the Einstein that will ship soon, using my design for an analog global power supply, and very sophisticated modeling lamp control and many other refinements. This effort took sixty days and was concluded last October.
I instructed the consultant to follow my analog core design to the letter and to simply do the pc board layouts, following my specific layout and package requirements, and to reprogram the ABMAx control uP board to the new requirements. A two month timeframe was allotted for this fairly straight forward work and a late 2009 schedule was established.
I have been pushing them almost daily since October to complete this work on the agreed scheduled and every week I have gotten board layouts with errors and incomplete programming. Dr. Morgan has been almost daily finding bugs introduced by the consultant and I have been doing the same.
As of last week, we had found and corrected all the bugs and felt we were a few days from doing the first per-production small run. I have been testing a completely packaged Einstein for over a month with excellent results. Most of the bugs have involved transient voltages when operating at 240V and the subsequent substitution of higher voltage/current components. (Our October prototypes used all through hole components and have worked flawlessly since they day we built it - with about a half million flashes on it.
In the conversion to surface mount components required for the final production the consultant did not do a great job of cross converting some of the through hole components to SMT parts. It goes like this: "Mike and Paul, here is this weeks's iteration - we have tested it it and it looks good." After Mike and I get our hands on it we find it doesn't work as well as they thought . . . this diode had to be changed, this resistor doesn't have sufficient power rating, etc."
To the credit of the consultant, they are sensation on digital and programming, but fall short on the analog and power circuit design that both Mike and I are expert in. This whole episode also proves that I am not expert at projecting the work of others outside my employee base.
So enough of this. The hopefully final circuit board designs are going to our Kansas robotic assembly partner this week. They will build a dozen or so and we will thoroughly test them and, if everything is as now indicated, we will be ready to produce the first run. I depend heavily on Dr Morgan for much of the bench work, and his unexpected seizure has put an additional burden on myself on on the rest of my technical staff
So there it its - the good, the bad and the ugly. I believe our very supportive customer base will accept these realities and I'm sure the few armchair critics will have a field day with judgements of me, of my company, my methods, my wide open communications, my marketing methods, my age, my red hair and anything else that brings them such joy.
Like it or not, I am not the typical corporation owner/executiveor CEO. I am a human being with a dedicated staff of 45 who believes in honesty, the Golden Rule and very open communications . . . and my company still has around 65% of the USA market. By the way, don't expect any runaway acceleration or braking problems and take my word I am doing everything I can to avoid any recalls or customer disappointments. To those who think I ought to act more like the CEO of IBM, this ain't IBM and proudly never will be.
Your turn - have a ball
|
It's a lot worse in terms of cost if a hardware bug gets out into the field (something like 10x the cost of a delay and fixing it before it gets shipped...I've been sent out to the field to hunt problems down), so it's better to do more testing, beta testing w/ guinea pigs, etc. to find them.
