Electronics > Repair

Looking for help fixing "dim" Advantest TR6847 multimeter

(1/1)

PlaneSailing:
I have an Advantest TR6847 meter which almost works nicely. It gives good measurements, but the 7 segment red LED displays are all dim. On the digital board the power to the display drivers and other digital chips is 4.7 Volts, as is the +5VL labeled test point.  I'm assuming the 4.7V supply is causing the dim LEDs.  The output of the LM7805 5V regulator though is good at 5.0 Volts, wondering what is going on ?.  ADCcorp reply to emails quickly, but there is no sharable service manual or schematic of the TR6845/6/7/8 series.

Powered off resistance between output of the LM7805 and +5VL test point is 10 kOhms and 1 kOhms with (a different) meter leads reversed, I expected a short.   If really 1 kOhms, a 0.3V drop implies only 0.3mA of current - way to low... something "funny" going on, so hesitant to simply connect the 5V regulator output and the +5VL test point with a wire.

Appreciate advice, like someone with a similar meter measuring the voltage at the digital board 5V regulator output, and the +5VL test point (both relative to TP1 GND), and resistance between 5V reg OP and +5VL TP.   A link to the schematic or service manual would be great, thanks much Peter

PlaneSailing:
Update: poked about and realized with this meter having the potential to be battery operated, the 5 Volt logic rail could be disconnected from the LM7805 regulator when battery power is in use - and the 6V ? battery instead supply the 5 V rail. An electronic disconnect could explain the 0.3V drop between the +5VL test point/logic rail, and regulator output. Summoning up courage, I shorted them together - no smoke, but also no increase in the 7 segment LED display brightness.

Digging in, found eight 220 Ohm resistors used as the current limits for the 7 segment (plus decimal point) displays, the scope showed a 2V drop across the 220 Ohm, so 9mA of LED drive current.  Paralleling the 220 Ohms in turn with another 220 Ohms noticeably brightened the display segments one by one.  Decided to replace the eight 220 Ohms with 100 Ohms, increasing the LED operating current from 9mA to 20mA. This made the display brighter, and 20mA should be OK for the LED current, especially at 1/8 duty cycle.  Maybe a low LED current is chosen by design to extend battery life ?.

Display still not as bright as my HP, Fluke or Solartron meters, but now very viewable, maybe Advantest 1980's LED multimeters are just that way ?.  An interesting worthwhile project for anyone with the same issue, possible even without a schematic.  The internal details and build quality of Japanese instruments and decent hi fi from this era are very good.

wn1fju:
I have both the TR6847 and the slightly less capable TR6845.  I opened them up and took some measurements:

On the TR6847, the output of the 5V regulator measures 5.6V with respect to TP1 (gnd).  And that is because, at least on my board, the ground of the 5V regulator IC is lifted up by 0.6V by a diode to ground.  I do not have a +5VL test point on the board.

On the TR6845, I get pretty much what you got.  Namely, 5V out of the regulator IC (there is no diode lifting the ground pin) and 4.77V at the +5VL test point.  The resistance measurements from the output of the regulator to the +5VL test point were 1K in one direction and moving (>10K) from some capacitor charging in the other direction.

I have attached a picture of both units in moderately dim ambient lighting.  I really can't tell any difference in brightness between the two units.  They are not extraordinarily bright in either case, but perfectly adequate in my opinion.

And no, I don't have the service manual either.  Would love to have one!

PlaneSailing:
Thanks wn1fju for your investigation, post and picture.
Interesting that Advantest lifted the 5V regulated output, suspect due to loosing voltage in the "using battery power" switch.  In your picture all the 7 segments have even brightness, this is also the case with pictures I have found on the web. For my unit, even after decreasing the LED current limit resistors to 100 Ohms, so increasing the LED current to 20mA, each segment of the display is brighter in the middle, tapering off at the ends where it meets the neighbor segment.  My unit is now usable - looks I just have "weak or dim" LEDS, interesting you describe your units as not bright but adequate - thanks.

Yes, a service manual/schematic would be wonderful.  The TR6845/6/7/8 family are nice meters, I notice your units measure the same voltage to within 5 digits.  They have some neat features compared to my similar HP3466A of the same era - though the HP has a bright display and available service manual & schematic !.
Hope this helps someone with the same issue.

Navigation

[0] Message Index