Stories from the Wall: Trojan UV3000 – changing the course of our business

I couldn’t have a wall-of-fame of “trophies” from my first company without having Trojan’s UV3000 system components in it. The UV3000 products were the catalyst opportunity for the company that changed the course of the company’s trajectory.

When two friends and I launched the company in 1980, it was a hobby sideline. I was the third to quit a full-time job and jumped full-time in 1981. We were a services company, fixing equipment made by others, and then adding some panel building and dabbling in other services we were able to provide. After 6 or 7 years working at it full-time, we had grown the company to about $750K per year in revenue.

Every month, we had to find some way to generate $60K to $70K in revenue, and our pipeline was barely 30 days out. We had a few good returning customers, but we still had to hope the phone would ring. Needless to say, we had quite a few lean months in those times.  

That all changed in 1988, with a letter of intent from Trojan to purchase about $750K of newly designed electronic assemblies over a 12-month period. That LOI commitment instantly doubled our annual projected revenue, and it hastened our transition to a company that manufactured its own product rather than being a services and build-to-order company.

But it took far more than a single event to get to that point. It wasn’t a “lightning strike”, as some like to put it.

In an earlier blog in this series, I wrote about the Hyd-Mech PLC100 being the first product the company designed where we didn’t charge for the design, designing it at no charge to the customer, ensuring we retained ownership of it.

We took the same approach with the UV3000 control system. Trojan had just completed the design of the main components of their then new UV3000 product family, before my company got involved. All the extensive testing in the lab had been done - but what works perfectly in the lab doesn’t always work as well in the field.

Their first system was installed in the Midwest, and their service people went on site to commission it. There they encountered unexpected problems with lamp modules failing to start, or ground-faulting off when other modules came on. After a few days, I was sent to the site to try to figure things out, and a few days later, I was able to get that first system stable enough to turn on and run.

Back in London the following week, I was invited to join their engineering team meeting to determine how to improve the system. There were 11 more systems ready to be shipped, and all the issues encountered in the first commissioning needed to be resolved. The central issue was that each module had eight lamps that were underwater, which required ground fault protection. The ballasts in each module were inductive loads, causing electrical spikes whenever they turned on. And the lamp monitoring and control circuits were largely analog and quite sensitive to electrical noise.

In the lab, with adequate clean power supply and all devices connected in close proximity, and only a small portion connected at any time, everything worked as intended. As was experienced in the field in the first installation, however, field conditions weren’t nearly as optimal.

Difficult (and courageous) decisions had to be made - and had to be made fast.    

After discussing all the ways to improve the current control design, the grounding of the major components, and strategies for dealing with less than optimal incoming power, they asked me for my opinion.

I recall my answer: “if it was up to me, I’d convert all the electronics to digital, communicate with isolated RS485, and eliminate all connections between system power and the control system”. (Sounds easy and perhaps obvious now, but it wasn’t in 1985).

They asked, “can you do that?”, to which I said I could. The VP then asked, “how long to build a prototype system for us to test?”

I said, “six weeks”.

It was audacious. Module boards had to be designed to read lamp status without connecting to the lamps or the ballasts. Relay boards had to be designed that had their own ground fault detection built in, and a circuit that ensured they only switched on or off at zero-cross.

Communication and control boards had to be designed for controlling the relay boards and communicating with all the module boards.

Each of these boards had to be laid out, produced, then rushed through production. Firmware had to be written for the module boards, and far more complicated firmware for the communication and control boards. Brand new communication protocols had to be created. Yes, indeed, it was audacious, nearly impossible. But we weren’t very busy, and we were desperate. This was an opportunity we had to jump through. So we did.   

“And if it works”, the VP replied, “how long to then go into production and build enough for

12 systems”. I gave him the same answer, “6 weeks”. (In for a penny, in for a pound!)

We had 6 weeks to build the prototype system for them to test, and if it worked and they approved it, we had 6 more weeks to finalize the designs and produce enough for 12 systems.  If it didn’t work, the future didn’t look good for either company.

The prototype system worked and we went into production to produce products for 12 more.

The order was then upgraded to 30 systems, but we weren’t out of the woods yet. There were a lot of early site visits, design modifications, and firmware development was still going on.

It was a stressful time, and being candid, very stressful within my own company. We weren’t charging for the design time, which included the continuing design refinement time, but we were now purchasing a lot of parts, paying a lot of set up and rush charges, and applying a lot of our people resources to it. I believed we had chosen the right path for the long term, but we had to survive the short term challenges to get there.

I met with Trojan’s President and explained the predicament. He conferred with their sales director to determine how many systems were likely to close in the year, and then with their purchasing manager. We then wrote up a LOI identifying all the assemblies they would purchase from us within a calendar year.

The total in the LOI was just over $750K, and they did in fact purchase that amount in 12 months. And each subsequent year for a good period of time, they purchased even more.

They took a chance on us, and some say we also took a chance on them. We certainly took a chance by incurring all the risk and cost designing the critical electronic control assemblies ourselves, but they supported this course, and it turned out well for our mutual benefit.

With the UV3000 opportunity, our company graduated from primarily being a service business and a contract manufacturer to a manufacturing company that designed and produced its own products. It was a catalyst opportunity, and a course change that has served the company well.

There will be many more stories from the wall to come, but in the mean time, here's one I wrote a while back that relates directly to this event.

Entrepreneurial Epiphanies: "the Company with the Best Customers, Wins."