|
I recently performed a few repairs on a friends 3457A multimeter (new RAM battery, replacement current range switching relays and added thermal insulation on the LM339 vreg) and have also just acquired one of my own that I'll do the RAM battery replacement, LM339 thermal insulation and maybe power supply capacitor replacement on.
I noticed there are a number of 51K carbon composition resistors in the signal path, which I found a bit strange...
You can see them in the input switching section highlighted in yellow of the linked image below, R101, R102, R201, R202, R203, R204
As carbon composition resistors are prone to drift and noise with age, would it be of any benefit to replace these with metal film resistors? Or is there some reason carbon composite were used here (high pulse energy handling and low inductance etc)?
Thanks!
Jared
|
composition resistors actually have more spurious reactance than film resistors. I have no idea why these were used but suspect carbon or metal film resistors would work at least as well.
toggle quoted message
Show quoted text
On 5/7/2024 5:36 PM, Jared Cabot via groups.io wrote: I recently performed a few repairs on a friends 3457A multimeter (new RAM battery, replacement current range switching relays and added thermal insulation on the LM339 vreg) and have also just acquired one of my own that I'll do the RAM battery replacement, LM339 thermal insulation and maybe power supply capacitor replacement on.
I noticed there are a number of 51K carbon composition resistors in the signal path, which I found a bit strange...
You can see them in the input switching section highlighted in yellow of the linked image below, R101, R102, R201, R202, R203, R204
<>
As carbon composition resistors are prone to drift and noise with age, would it be of any benefit to replace these with metal film resistors? Or is there some reason carbon composite were used here (high pulse energy handling and low inductance etc)?
Thanks!
Jared
--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998
|
a lot of times carbon
comps were used as they could handle surge I better and not go
open
搁别苍é别
On 5/7/24 6:08 PM, Richard Knoppow via
groups.io wrote:
toggle quoted message
Show quoted text
??
composition resistors actually have more spurious reactance than
film resistors. I have no idea why these were used but suspect
carbon or metal film resistors would work at least as well.
On 5/7/2024 5:36 PM, Jared Cabot via groups.io wrote:
I recently performed a few repairs on a
friends 3457A multimeter (new RAM battery, replacement current
range switching relays and added thermal insulation on the LM339
vreg) and have also just acquired one of my own that I'll do the
RAM battery replacement, LM339 thermal insulation and maybe
power supply capacitor replacement on.
I noticed there are a number of 51K carbon composition resistors
in the signal path, which I found a bit strange...
You can see them in the input switching section highlighted in
yellow of the linked image below, R101, R102, R201, R202, R203,
R204
As carbon composition resistors are prone to drift and noise
with age, would it be of any benefit to replace these with metal
film resistors? Or is there some reason carbon composite were
used here (high pulse energy handling and low inductance etc)?
Thanks!
Jared
|
These resistors look like they are part of an overload protection circuit in conjunction with the clamping diodes in U101.? Carbon composition resistors are used here for their much better surge handling capacity than film type resistors. I wouldn't replace them with film resistors if you want to retain the best input protection for the (probably unobtainum) hybrid amplifier module.
The drift of a carbon composition resistors would have negligible effect on the accuracy of the circuit as they are in series with some very high impedance amplifier inputs.
|
There's no need to change those 51 k resistors. HP had it all figured out, for proper and safe input overload handling and ESD. You'll probably notice they are fairly large (maybe 1 or 2W) in power rating. That comes into play in the event of sustained over-voltage on the input connection, where the voltage clamps in the front-end limit the circuit voltage, and the resistors limit the current and take the rest of the total voltage, and have to dissipate the power. To appreciate the size needed, just say, take the specified maximum input overload voltage squared over the total R, for total power. Normally, these parts seldom need to do their thing, but they're always ready. If they appear overheated or burnt, then the unit may have been subject to severe abuse.
Ed
|
Thanks for the replies.
?
I am inclined to leave the resistors alone, but I just keep wondering how much their temperature coefficient and value drift with age would be affecting the accuracy of the instrument?
Against my Fluke 731B volt standard connected with Pomona low EMF cables, the first 3457A meter I repaired exhibited a drift due to temperature from 10.00016V to 10.00020V between 23 to 27 degrees C.
The second 3457A should arrive today or tomorrow so I'll be able to test it after I have had it powered on for a day or two to settle.
On my 34461A under the same test conditions, the voltage range was 9.99995V to 9.99996V
Jared
|
Jared, A couple of points: - Please make sure you account for your 731B's own tempco, which typically is not negligible. I have one that's unusually?stable, but most units exhibit quite a substantial delta over temperature?(for a voltage standard). Certainly, seeing the .25ppm/C on your 34461A seems to point to the tempco occurring?in the 3457A. When looking at this, though I try to account for counterproductive synergies - is my voltage standard based on the same internal voltage reference as the meter? Etc.?
- I think Stephen makes a pertinent point with the impedance of the input circuitry. I'm not sure you should see much influence from those input resistors, given they look like shorts to the input of the meter. I'd suspect the TC you're seeing may be due to other effects and parts of the circuit, and likely much harder to cure.?
Also, I couldn't find this data, but 1ppm/C doesn't strike me as excessive from this instrument (particularly?given that the 3456A seems to have a TC of about 2ppm/C).?
Now, if not too difficult to work on, you could try replacing them with some good types - maybe something like CMF60s or even Mills WWs - and see if you observe an improvement of performance.? Radu.?
Thanks for the replies.
?
I am inclined to leave the resistors alone, but I just keep wondering how much their temperature coefficient and value drift with age would be affecting the accuracy of the instrument?
Against my Fluke 731B volt standard connected with Pomona low EMF cables, the first 3457A meter I repaired exhibited a drift due to temperature from 10.00016V to 10.00020V between 23 to 27 degrees C.
The second 3457A should arrive today or tomorrow so I'll be able to test it after I have had it powered on for a day or two to settle.
On my 34461A under the same test conditions, the voltage range was 9.99995V to 9.99996V
Jared
|
The carbon composition resistors were probably used with the thought that they are rugged, and will only introduce a variation that is less than system noise. However, carbon composition resistors that are old, and subject to humidity do strange things. Those over 100K seem to always go down in value, and those below 100K go up in value. I don't know why, but they do. Old carbon composition resistors typically move much more than their tolerance would suggest... very much more..., which may push their effect from negligible into very noticeable. In the case when the resistor has been left in an unused state over a long time, in a device, when it is first powered on, and for many hours(days) thereafter, it will drift.... But, only in one direction. As an example, I have a 432B power meter that took 2 months of being powered on 24/7 before its drift slowed to an almost stop. -Chuck Harris On Wed, 08 May 2024 16:10:50 -0700 "Jared Cabot via groups.io" <jaredcabot@...> wrote: Thanks for the replies.
I am inclined to leave the resistors alone, but I just keep wondering
how much their temperature coefficient and value drift with age would
be affecting the accuracy of the instrument?
Against my Fluke 731B volt standard connected with Pomona low EMF
cables, the first 3457A meter I repaired exhibited a drift due to
temperature from 10.00016V to 10.00020V between 23 to 27 degrees C.
The second 3457A should arrive today or tomorrow so I'll be able to
test it after I have had it powered on for a day or two to settle. On
my 34461A under the same test conditions, the voltage range was
9.99995V to 9.99996V
Jared
|
The resistors are in a high impedance part of the circuit so the absolute value will not affect the accuracy of the readings.
As others have said, part of the input protection and probably chosen for surage capability, power and voltage handling.? Note the used two in series. This will be to increase the power and / or voltage rating. Unless thay are well out of specification (>10%) I'd leave well alone.
Robert.
|
Well, I lifted one leg of each resistor, let them cool to ambient and measured the resistance.
Here's what I got. (They are 5% resistors).
55.20K - 7.91%
54.41K - 6.51%
56.15K - 9.61%
53.67K - 5.10%
55.93K - 9.22%
53.36K - 4.52%
As an aside, this will be getting sent off for cal/adjustment soon too (probably to Keysight if they still cal these) and I have no cal history on the unit, so I have no concerns about the calibration being off after I poke things.
If I were to replace these, I have these in my stock of parts:
RLC50 series - 51K 1% 50ppm/C, 350V working, 700V overload, 0.3μV/V max current noise.
Or I could get something better from Digikey/Mouser.
Or I could just leave them alone and deal with the effects of any drift...
Jared
|
Jared, The drift you're seeing is not excessive from that type (and age) of resistor. You could still replace them with brand new ones (the "for" for that would be long term?service and peace of mind - they'll only continue drifting from this point on, though regularly running a long-stored unit will help those resistors, as well as all parts, do better), you'd just have to be mindful of their role when choosing a type or resistor (per above points). I'd do the replacement, obviously, before sending the units in for calibration.? I checked on the US portal for Keysight services (which is where I generate quotations, etc.) and, for some reason, I do see the 3456A, I certainly?see the 3458A and also the 34401A (the latter two I know for a fact they are supported), but I don't see the 3457A. I'd be curious if they support it for cal where you are.? Radu.?
Well, I lifted one leg of each resistor, let them cool to ambient and measured the resistance.
Here's what I got. (They are 5% resistors).
55.20K - 7.91%
54.41K - 6.51%
56.15K - 9.61%
53.67K - 5.10%
55.93K - 9.22%
53.36K - 4.52%
As an aside, this will be getting sent off for cal/adjustment soon too (probably to Keysight if they still cal these) and I have no cal history on the unit, so I have no concerns about the calibration being off after I poke things.
If I were to replace these, I have these in my stock of parts:
RLC50 series - 51K 1% 50ppm/C, 350V working, 700V overload, 0.3μV/V max current noise.
Or I could get something better from Digikey/Mouser.
Or I could just leave them alone and deal with the effects of any drift...
Jared
|
I know from personal experience that
the ceramics and film under surge current will
typically be destroyed ( almost like RIFA caps just not as
messy) ..hence? the Carbon comps...
搁别苍é别
On 5/9/24 6:49 AM, Jared Cabot via
groups.io wrote:
toggle quoted message
Show quoted text
Well, I lifted one leg of each resistor, let them cool to
ambient and measured the resistance.
Here's what I got. (They are 5% resistors).
55.20K - 7.91%
54.41K - 6.51%
56.15K - 9.61%
53.67K - 5.10%
55.93K - 9.22%
53.36K - 4.52%
As an aside, this will be getting sent off for cal/adjustment
soon too (probably to Keysight if they still cal these) and I
have no cal history on the unit, so I have no concerns about the
calibration being off after I poke things.
If I were to replace these, I have these in my stock of parts:
RLC50 series - 51K 1% 50ppm/C, 350V working, 700V overload,
0.3μV/V max current noise.
Or I could get something better from Digikey/Mouser.
Or I could just leave them alone and deal with the effects of
any drift...
Jared
|
The drift in the resistor values doesn't look too bad as it is about what you expect for carbon composition resistors of that age so I wouldn't replace them.
The small drift in values is going to have negligible impact of the accuracy of the meter as you will only get the input leakage current, probably a few pA, of the gates of the input FETs and their associated protection diodes flowing through them. The error contribution from the drift in these resistors is therefore only a few nanovolts and will barely nudge the LSD of the readings on the most sensitive range. You are likely to get far greater errors from thermal voltages in your test leads.
I specifically wouldn't replace the carbon composition resistors with your suggested metal film types as these are likely to have significantly poorer surge overload performance than the original carbon composition types. If you really must replace these resistors you should use the original carbon composition types, if you can find them, or the more modern ceramic composition types, if you can afford them.
|
Having a closer look at the boards, I've noticed that the resistors are all on the various inputs, there are two more carbon composition resistors between the front binding posts and hybrid (for other functions and ranges) and also on the trigger input from the respective BNC on the rear panel. There is also one across the control data lines to the optional relay card too, so I think it is likely that they were selected for surge rating to handle any transients that may come in.
The True RMS converter board also has a few, but two are 100Meg resistors, which I believe would have been chosen due to production limitations of the time making such high value resistors unavailable in 1/4 watt size metal and carbon film resistors.
As for calibration, I assumed this would be able to get done as the other models are still serviced by Keysight. For the 3457A, they just say to contact them, so I will have to do so and see what they say....
Jared
|
Jared Cabot