After having assembled the Cyrob CP0096 differential probe, which is electronics enthusiast designed, I wanted to have a close look into a more professionnal differential probe: The TA041 from Pico Technology. A look so close that I'll re-draw the schematics of that thing. It is a ÷10 or ÷100 probe, running on 4 AA batteries (6V total) or 6 or 9V external voltage.


The TA041 probe, without its yellow jacket


Output BNC cable is going thru batteries (also, nice hand-written serial label)


Once the plastic case opened, we can see two separate shielding enclosures. The left one is containing the analog circuits, and has multiples holes for adjusting. Not all holes are having an adjustable pot or cap behind. Maybe used for several probe versions?


The front plastic shell halve is also having a good number of holes. Only one is not covered by the adhesive plastic sheet and allowed to be touched by the end user (if we exclude the ones for switchs and led of course).


Printed circuit board bottom view, shields removed


Printed circuit board top view, shields removed. Input divider in two boards, and three ICs are painted (but for what purpose?)

LabelIC before paint removalAfter paint removalMarkingsWhat this thing probably is
U1LS591B F
1304 C
Wideband, High Gain, Monolithic Dual, N-Channel JFET
U2JM25ABE3
LM3046M
NPN Bipolar Monolithic Transistor Array
U3MKAB
????
74A IM
Op-amp
U4 & U5TAC3Programmable Zener
(U4 not populated)
U6390V
S18B
Step-up PWM DC/DC Converter
U7LM393
39K
C2Q5 S4
Low Offset Voltage Dual Comparators


Top, with input resistances boards removed (capacitive divider exposed) and ICs paint cleaned.


Reversing the boards, both sides...


Reversed schematics. Click to enlarge!

I had to desolder a HV input capacitor from its string to measure it as a 10pF. After the divider, for each input: ~330pF//26.2kΩ to ground, measured at 10kHz. There is a lot of strange squirks in that device, like two components placed on the same SMD footprint (for both serials or parallels combinations), or replacements (diode mounted in place of an inductor, resistor in place of a capacitor, solders points in place of resistors... etc.). And, of course, some components just aren't populated.

LabelsMarkingValueLabelsMarkingValue
R1..R204023402kΩR45,R46(not populated)(infinite)
R21,R22(NM)457ΩR47,R4868B4.99kΩ
R23,R261501 + (NM)53.86kΩR49
R24(NM)R50
R25(NM)52.13kΩR5122322kΩ
R27(NM) // ?R52,R5939C24.9kΩ
R28(NM)R53
R29,R3384A732ΩR54,R55,R74,R75(solder point)(zero)
R30(NM)R56
R31(NM)R5711C12.7kΩ
R32(NM)R5833333kΩ
R3424B1.74kΩR60(NM) // ?
R35,R3735A226ΩR61,R6367C48.7kΩ
R366226.2kΩR62,R65,R72,R7341C26.1kΩ
R38? // CR64,R6639D249kΩ
R3984B // ?7.32k // ?R6769C51.1kΩ
R4068X49.9ΩR6861C42.2kΩ
R41R6938B2.43kΩ
R42R7016C14.3kΩ
R4385B7.5kΩR7101D100kΩ
R4449R949.9ΩC3495B9.53kΩ

Resistors values. Not all. Most aren't marked and I desoldered only some for measurements...


1.5k resistor added in serie with 52.3k (for fine tuning?). This is the kind of thing you don't see everyday in a marketed product...


Closeup on the input divider section, low sides. One capacitor is mounted on top of another and not on its own footprint. But why?


SOT23 diode mounted in place of an inductor named L4