[Premium-Rx] History of the WJ-8711

[spencer]

Terry,
I really appreciate when you share your historical and programmatic information with the group. It’s great stuff! Thank you.
Spencer

> On Nov 17, 2018, at 3:39 PM, Terry O' <watkins-johnson at terryo.org> wrote:
> 
> Rick Lober was the manager of the Gaithersburg facility during the development of the WJ-8711.  For those of you who don't use (or have a low opinion of) Facebook, below is what he just posted in the BlackRadios group:
> 
> Best,
> Terry O'
> 
> About 1990, one of WJ's larger customers asked if a receiver could be designed with the high level of performance found in the popular WJ-8718 HF Receiver (selling for about $15,000-18,000 at the time) but near the cost of the Japanese ICOM units they were buying but were not happy with. They set a goal of $3000 and told us they would buy 1000s if WJ could pull this off.
> 
> Bill Bruff, who was division manager at the time, agreed to fund the product with R&D. The customer placed an $18,000 order for 6 production units at $3,000 each (no NRE...) with a delivery time of 9 months (a very short development period for a project like this). Novel "like the book" Fullen was the sales lead who booked this whopping first order.
> 
> Steve Hedges was selected as Project Engineer and put together a great team. I led the overall program which included working with purchasing and manufacturing to help in hitting the cost target while Steve's team knocked out the design.
> 
> Requirements definition was key and once set, the team worked hard to avoid feature creep. The only way to hit the cost goal was to use a DSP based IF section - a first for WJ. Rockwell-Collins (of R-390 fame) did release a DSP HF Receiver about 1-2 years ahead of us but the unit drew lots of power, was expensive and had heat issues. Two years later, technology had advanced and the time was right...
> 
> The design team settled on a single board radio that included the RF section and A/D converter (a 1 bit Sigma-Delta type) followed by a single Motorola 56000 DSP chip that would perform all the functions usually done by analog IF circuitry in the conventional receivers of the time. The HF preselector was an option (as it was in the 8718) and the rest of the unit included a switching power supply and front panel board.
> 
> Extensive time was put into the front panel layout and control software to insure it was intuitive and included all memory and scanning features along with the infamous WJ large tuning knob that the customer insisted on. Bob Watson analyzed all of that in his usual detailed manner.
> 
> Using DSP, the team was able to achieve a narrow band IF filter of 100 Hz - a first and great for pulling CW out of the noise. Tuning step size was as low as 1 HZ - another first - all due to the use of Digital Signal Processing in the IF section (sometimes called software defined radio now).
> 
> About 50% of the way thru the program, I asked Steve what the DSP utilization was - he stated it was nearing 90% - I about panicked and asked why not add another DSP to share the load. Steve and team assured me they could tighten up the code, that the cost goal was paramount, and adding another $100 chip was not going to happen - they were a determined bunch!
> 
> We had the same level of commitment to that goal in purchasing and in manufacturing - at times the staff was trying to save pennies and between the DSP loading issue and everyone doing backflips to save a cent I wondered if we would miss the forest for the trees and end up with nothing. In the end, engineering pulled it off as did the purchasing and manufacturing team.
> 
> The night before the customer was to arrive for sell-off of the six units, the team went late into the night. I decided to put my engineer/former HAM hat back on and play with one of the units using some real signals. At about 11 PM, I found that the USB/LSB sidebands were reversed (or maybe they were mirrored...). Steve grabbed Al Blount who was the lead DSP engineer on the project and the code was corrected. The team went home about 6 AM, showered, put on ties and showed up at 9 to meet the customer. 3 hours later the 6 units were sold off with a few minor issues. Steve and team had pulled it off in 9 months and I was honored to be involved in a "once in a career" effort.
> 
> In the end, 1000s were sold and other configurations were designed including a half rack unit with front panel and a blank panel computer controlled unit.
> 
> In addition to our primary customer, 100s of the half rack version were also used by the BBC in London for the "BBC World Service" - a shortwave monitoring system that transcribed international broadcasts. They wanted the best that was out there for this important service.
> 
> We also entered the Shortwave Listeners/HAM radio market with the unit and called it the HF-1000 - - but that's a story for another day...
> 
> 
> 
> ______________________________________________________________
> Premium-Rx mailing list
> Home: http://mailman.qth.net/mailman/listinfo/premium-rx
> Help Page: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Premium-Rx at mailman.qth.net
> Help Contact eMail:  radio at 8zo.com
> Home Page:  http://www.premium-rx.org/