Ads skipped

EEVblog #1189 - Ebay $140 6.5 digit DMM Bargain!?

172K views · Mar 12, 2019 · Science & Technology

EEVblog

1M subscribers

Comments · 445

@EEVblog · 7 years ago · pinned
Got an email from a viewer without a Youtube account: Quote: "The reference in your < US$200.00, 6.5 digit multimeter blog has some really ‘old school’ wizardry in it. Way cool! If you are interested, I would like to point out a few things lurking inside the LM136: First its (surprisingly) mostly a Brokaw bandgap cell. Being National Semiconductor, one would have expected it to be a Widler bandgap cell IF it went all the way back to the birth of the bandgap reference, which happened at National. The Brokaw part of it uses 2 transistors (versus 3 for the Widler) in Q15 & Q16, which are almost certainly laid out with an emitter area ratio of 8:1. If both Qs are operating at the same current, the voltage across resistor R8 will be kT/q * Ln(8), or about 53.44 mV. At room temperature. This is multiplied up via R7, R8 and R10 by 23X, which is about 1.2292 Vdc. This is added to Q17 & Q18 Vbe connected diodes. Assume Vbe = 0.6354 Vdc and you get a total of 2.5000 Vdc! The collectors of Q15 & Q16 have equal load resistors and Q7 & Q8 are the input stage of the op-amp that balances the Brokaw cell. Q1 is the ‘power’ shunt pull down regulator. Actually, this is not quite a Brokaw cell (due to patent concerns), so the tail current of the first stage is NOT derived by reflecting the collector currents back down to the current source that feels them. Instead, a really neat hardly ever talked about ‘Cross-Coupled Cell’ is used to generate a PTAT (Proportional To Absolute Temperature) current at emitter-degenerated Q14. The flatness of a bandgap over temperature is much better when the collector currents are PTAT than when they are NPO. (Note: the ideal TC is process dependent on the so-called ‘eta’ factor which, if close to 2.0, all is golden). This CCC (cross-coupled cell = Q10, Q11, Q12 & Q13) has a difference in ‘bottom’ emitter voltages that is independent of the ratio of left to right collector currents! This happens because the Vbe(Ic) for both Left & right show up in the sum of Vbes on BOTH SIDES! This is so useful, Dave. For example, if you need a really low input or termination impedance and you don’t want to throw current and power at it, this is the cat’s meow. Zin can be Beta/2 lower than Re (emitter impedance). For example, at 1 mA, Re = 25.7 Ohms, but the CCC can have an input impedance of 0.1 to 0.25 Ohms at the same current. Here, those bottom NPNs also have an area ratio (probably 8:1 also) and this makes a PTAT voltage across resistor R2, and that flows into a, ready for it? A Split-Collector lateral PNP! Q9. That produces an equal PTAT current copy that feeds the tail of the lateral PNP 1st stage transistors, Q7 & Q8. Underneath is the NPN current mirror, Q4 and Q6. The next neat trick is that the sum of the current mirror transistors’ current is also PTAT, and that is used to feed another NPN current mirror Q3 & Q5 whose job it is to make the collector, and therefore the base currents of 2nd stage NPNs Q4 & Q6 not only equal but also PTAT. Making everything possible run at PTAT currents keeps all sources of errors essentially constant over temperature, and is almost always exploited. Finally, Q1 is the final stage of the balancing op-amp, and it operates in the shunt mode, diverting any current in excess of the sum of all the other transistor currents, all of which are running PTAT. R1 is a ‘pinch-resistor,’ which gives very high resistance in limited areas but is otherwise abysmal, with wildly varying initial value and horrible TCs. It is used as a fail safe startup circuit and the only reason you can get away with using it is because of the in
69
@excavatoree · 7 years ago
The "Dave Effect" will make it impossible to duplicate this kind of a bargain for a while.
175
@trevorvanbremen4718 · 7 years ago
And suddenly, the EBAY pricing of Philips Bench DMMs goes thru the roof!
30
@Biela2008 · 7 years ago
Backlit LED display mod next! :D
61
@Rob2 · 7 years ago (edited)
Philips - Made in Holland All those original Philips components... brings a tear to the eye.
87
@ChristopherJohnsons · 7 years ago
I recently read a story about a bike manufacturer from the Netherlands that had massive problems in the US with damaged goods due to shipping. Their solution to the problem was to print a TV on the outside of the package because TVs are expected to be more important than bikes in the US ;) Their returns due to damaged shipments dropped by 80 percent right after that. So you clearly should give any seller from the US the hint to mark the package with "TV inside" or the like to have it shipped without damage :D
215
@Henchman1977 · 7 years ago
I sell on ebay for a living (mostly used items). There's absolutely no excuse for that packing job.
79
@kippie80 · 7 years ago
Dave, dont forget to re-label the back. She's 230V now!
54
@BlackEpyon · 7 years ago
I've gotten stuff from China that was better packaged than that. Yikes! All things considered though, not a bad score!
28
@thermionic1234567 · 7 years ago
Actually it’s “US$150 just-barely delivered to Australia!”
8
@ckm-mkc · 7 years ago (edited)
Backlight is burnt out.  It's a EL strip and is a pain to replace....  You can get a larger EL strip & cut it down, but IRC it's a little involved (very tight fit) but only the EL strip needs replacing, the HV power supply is fine.  BTDT on a similar display but on a power supply.  That old Philips stuff is great quality overall, except the backlights...
12
@kefler187 · 6 years ago
I think the panels are designed to be swap-able depending on where you want your jacks. If you want rear facing jacks for continuous monitoring through GPIB then you can swap the clean panel to the front and the jack panel to the rear.
4