Any Questions ?
Call Us
(508) 881-99
59
M-F 10am - 6pm Eastern

 


Hi-Rez Projections
20 Main Street
Ashland, Ma 01721
Office (508) 881-8859
Fax (508) 881-9958
April 13, 2005


Marquee Low Voltage Power Supply Fan and Heater Voltage circuit modifications
While working on Electrohome Marquee chassis over the past eight years, I’ve found several areas that would benefit, both in performance and reliability, from modifications to the Low Voltage Power Supply (LVPS). This paper discusses the resulting modifications to the LVPS, and their benefits.

Heater voltage circuit modification:

In 1998, Electrohome issued a service bulletin for the LVPS that addressed an issue the company was having with heater voltage runaway in units produced from August 1996 through August 1998. Due to a bad batch of variable resistors (P14), the LVPS heater circuit output could go high (above 7volts). The immediate result would be just a slightly brighter output from the tubes, but more importantly, this over-voltage to the CRT filaments would quickly lead to damage to all three tubes. After completing several costly repairs for this issue, I decided to take a look at the heater supply circuit, and see if it could be improved upon for reliability.

The original heater circuit uses a variable-voltage regulator (LM 317) whose output is set with a voltage divider network. Since there was a fair amount of variance within the LM317s, a trim potentiometer was designed into the circuit to make the final voltage output adjustment on each unit. Electrohome engineers chose 6.35 volts as the optimum heater voltage setting, which was also marked on the circuit board, next to the trim pot.

While examining the heater circuit, I also found that noise from the heater-voltage supply was being injected into the front-end video amplifier stage of the neck board. Pre-stage and post-stage filtering helped lower the noise somewhat, but by utilizing FFT analysis, it was realized that noise from the voltage-regulator IC itself was the primary cause of noise injected into the neck boards. As a result, I chose a new, low-noise variable-voltage regulator to replace the LM317. Additional circuitry was also added to lower the dissipation and operating temperature of the new regulator, which helps keep its noise level down and extend its long-term reliability.

In addition to its lower noise level, the new regulator IC has substantially better voltage setting accuracy and regulation than the LM317. This allowed the replacement of the troublesome variable resistor (P14) with a fixed network of 1% metal oxide resistors. Measurements on the first 20 modified power supplies showed a heater voltage output accuracy of better than 1%. In addition, the input and output filtering was modified to further lower noise levels in the heater output circuit and allow better output transient response.

In addition to these modifications, a new clamping circuit was integrated into the output section of the heater supply. In the unlikely event that the voltage regulator should degrade or fail and its output rise above it’s set level, the output voltage to the tube heaters will be clamped to a safe operating level.

The end result of these electronic modifications of the LVPS heater supply circuit is better and more reliable regulation, with no chance of drift or misalignment, a fail-safe clamp, and less video noise injection into the video circuits on the neck boards.

Power supply Cooling Modification
The Electrohome Marquee projector is designed to meet military specifications and work in tough environments, with 24-hour-per-day operation in mind. To cover worst-case conditions, the cooling system for the projector was over designed with a secondary regard to noise levels. This noise may not be important in situation, simulation, or boardroom applications, but is very important in home theater applications.

To cool the low voltage power supply, Spellman, the OEM of this supply for Electrohome, utilized two high-speed DC 3.5” fans. When testing sound levels around the low voltage power supply, I found that the most disturbing part of the fan noise was caused by acoustic nodal and anti-nodal points between the rotating fan blades due to their not being in sync. This resulted in a beating effect that is much more noticeable audibly than the steady noise of a single fan. To correct for this issue, an external fan plenum was designed to support a single, larger and lower speed 4.5” blade fan.

The new fan/ plenum combination had two positive effects on performance over the original design. First, it completely removes the beating noise of the two-fan system and is quieter overall as well. Internal thermal measurements to 20 tested units yielded the same, or slightly lower, internal temperature as the two-fan design.

Second, the new fan/plenum combination closes most of the air gap between the high voltage and the low voltage power supply so that it draws some air through the HVPS. With the original design, there is a 5.5” air gap between the low voltage and the high voltage power supplies. Unlike the low voltage power supply, the high voltage power supply relies entirely on convection cooling. By thermally coupling these two supplies, the internal temperatures of the high voltage power supply are measurably lower than in the original configuration. This in turn means longer life to one of the most expensive components in the projector.

The end result of this new LVPS cooling design is a measurably quieter (4.8 dB, A weighted) projector with enhanced cooling capabilities for both power supplies.


Steven Flynn
Engineering/ Technical services
Hi-Rez Projections, Crystal View Inc.