HP3458A FLATNESS DAC CONVERGENCE 195 error

[Kleinstein]

The schematics (from TiN's website) has some missing parts in where the ref signal is comming from.  It looks like there are 2 passible paths to choose from.
One looks like a 1/10 divider (R202,R201 = 9K and 1 K)  from the guard signal and thus 1/10 of the input to the amplifier. This would also be relatively small: 1/200 or 1/2000 of the amplifier output.

The lower ranges (10 mV,100mV and 1 V) could also use the gain 0.001 signal and thus 1/1000 the input signal and thus also a rather low level.

There likely is a reason to use such a small signal and than have a gain of 10 after the DAC.

[Traceless]

Thanks traceless,

The bit captured in the following was by single shot triggering on +9V rising edge. At approx t=6s after triggering, the beep indicating the error occurs.

There are +12 / -15v peaks then +3/-3v levels (approx 31 of each) - with spikes of varying amplitude. Not sure if this is revealing anything useful.

I have seen someone else refer to an issue with Q801/2 elsewhere, can't think where. I do have a couple of spares so it may be worth checking them as you say

I'm not sure, the peaks you recorded are extremely short and a lot less frequent compared to the case I had in mind (see video link below, the interesting part for you starts at around 25 minutes into the video). Those transient spikes may or may not be okay.

- I believe that part of the circuitry is generating the cal voltage input?

Yes you can see it on this schematic* the CAL line from the "Calibration" part of the circuit where Q801 and Q802 are located (page 3) is connected to the input of K203 which is part of the "High Voltage Attenuators and Input Switching circuit" on page 1.


(*While the schematic linked above is a lot easier to read than this one Q802 seems to be mislabled as Q803 compared to the original documents (p. 53)).

[chilternview]

Finally a AD7420CJ DAC arrived in the post. With the new part I eagerly awaited a self test... and it failed again with the same error.

[seebeck]

Hi Folks,
My 3458A turned out to have the same problem ，“Self test failed” --> "Flatness DAC Convergence:198" and showed 3xxx.xxxV on ACV with shorted input such as MiDi's plot# https://www.eevblog.com/forum/metrology/3458a-worklog/msg3627094/#msg3627094.
According to https://xdevs.com/fix/hp3458a_u4/, measured TP401 -7.89 VDC, then to checked U404 EL2039, inverting input had +6.892 VDC, non-inverting input +7.256 VDC.
I have been removed U404 and measured the pin1 non-inverting input and pin14 inverting input, voltages are 7.258 VDC and 7.093 V. Normally，TP401 should be equal to ~1.65 VDC if U404 is working well. It's a common issue for old 3458A.
U404 Elantec EL2039 is EOL, and Elantec as been bought up, Intersil... HA-2839（ref to https://www.alldatasheetcn.com/datasheet-pdf/pdf/91712/INTERSIL/HA-2839.html） is a p2p for EL2039. Currently, Renesas Electronics. neither of them had a great substitute for it.
EL2039/HA-2039/HA-2839，600MHz GBW and 600V/us high slew rate.. In the open market，you can't get any good one these days.
Q: How to replace this IC? Do you have any suggestions?

BR.
Seebeck

[syau]

Possible reliable source for EL2039 is utsource

[seebeck]

Hi Dr. Frank,

I found ADI, NXP and Signetics ever had a p2p device and their basic spec are almost same as EL2039.
The model is ADI AD5539 / NXP(early produced by Philips) or Signetics (currently, TI) NE5539, but this device are also being discontinued. 
AD5539's TDR is provided in attachment, please have a review.
Thanks a lot!

BR.
Seebeck

[Dr. Frank]

Supply voltage is +/- 10V only, so this seems to be the culprit, same goes for the other ELANTEC comparators.
You should have a look on the PCBs, how these parts have been replaced in the 3458A Black Edition.

Somebody has high-res photos on his site 

For the fast comparators, +/-5V types from T.I. I tink were used, and individual +/-5V supplies with small 78/79L05 were added.

Frank

[seebeck]

Well, really!
EL2039's supply voltage is +/-15 VDC, 10V is too low. AD5539/NE5539 was mainly used in audio circuit.

Somebody I'll check it immediately.
Thanks a lot.

With best regards.
Seebeck.

[Kleinstein]

The clipping detection is at +-10 V for the output. One might get away with a slightly lower supply, as long as the amplifier can drive up to +-10 V. So from the supply a AD8021 (+-12 V max) may be boarderline working, though a little slower.

The new black edition may give a hint on what they use.

A different amplifier may want some extra supply regulation and possibly slightly different capacitors for the compensation. The higher frequency end of interest can already be effected by the input capacitance and details of the frequency response. There are more parameters than just the GBW. Just the amplifiaction would get away with a little less.