Agilent 34410A, Repair of 2 dead meters [success]

[HighVoltage]

Hello Frank,

Ok I now measured the resistance and voltages applied to the heating element of the 10 M divider

Resistance: 54.87 Ohm
Voltage applied: 2.33 Volts

This is after warming up the instrument.
The other repaired instrument was cold and the resistor is 54.77 Ohm

[Dr. Frank]

HighVoltage,
thank you very much!

I also opened my 34465A, and they still use exactly the same TaN divider from IRC, and the same technique. I found 55 Ohm and 2,311V.

The trick is to compensate the self heating at 100V .. 1000V input voltages by applying external heat by this 55Ohm resistor.
(2,33 V over 54,9 Ohm gives 99mW, like in my instrument)

In any case, the total power dissipation is always about 100mW, so that the temperature of the ceramic array is constant in first order. (see attached table)
Changing the ambient temperature might be a 2nd order effect only.

So the ratio of this divider is always constant.

Without this heating, the ratio changes by +10ppm.


I always wondered, why the 3441x, 3446x and 34470 have no special HV calibration any more, only 500V.
The 34401A, instead, has to be calibrated at 1000V and once at 500V to compensate numerically for this effect.
The 3458A has no self heating compensation at all, and might drift by max. 12ppm @ 1000V. (my 3458A performs better, drifting only -1 .. -2ppm)

P.S.: During the measurement, I discovered a new bug in FW 2.10 (not yet released officially), which concerns this heating mechanism.

Frank

[HighVoltage]

Hello Frank,

Thanks for the explanation.
Good to know that this is the same technology in the newer meters.
I will take some IR pictures with my FLIR in the next few days and measure the surface temperature of the R225.
By touch I would guess around 40 degree C

[Vgkid]

That is pretty interesting how they used a heater compensated voltage divider resistor.

[HighVoltage]

It looks like other people have searched for relay alternatives as well and have not been really successful

Help me find a replacement relay...
https://www.eevblog.com/forum/chat/help-me-find-a-replacement-relay/

I will post close up pictures here as well.
May be some day, someone will find a replacement and we don't have to spend this outrageous amount of money at Keysight.

[HighVoltage]

@Frank,

I just measured the voltage at the heating element of R225 during power-up
of the 34410A. For a few seconds, the voltage sits at 3.16V, before it drops
to 2.33 V.
See picture

There is also a slight difference in voltage at the heating element, depending
on the range selection, but it is really only visible in the uV range of the trend
chart of the 34461A. (It shows the power of the trend chart in the 34461A)



 

[branadic]

It looks like other people have searched for relay alternatives as well and have not been really successful

Help me find a replacement relay...
https://www.eevblog.com/forum/chat/help-me-find-a-replacement-relay/

I will post close up pictures here as well.
May be some day, someone will find a replacement and we don't have to spend this outrageous amount of money at Keysight.

Have you tried contacting Meder? In the past they have been very helpful to me with their HF/RF reed relays. Maybe they help you too, as their slogan is "We matter.". You can refer to this when writing them 

[Macbeth]

I will post close up pictures here as well.
May be some day, someone will find a replacement and we don't have to spend this outrageous amount of money at Keysight.

No pictures of its internals? I always tear these things apart when they are done for. Sometimes you can get lucky and fix a mechanical problem and get some extra life out of them.

[HighVoltage]

No pictures of its internals? I always tear these things apart when they are done for. Sometimes you can get lucky and fix a mechanical problem and get some extra life out of them.

Well, I thought about taking it apart, but ....  it is potted with epoxy resin.
I will try to remove the epoxy slowly with some acetone.
May be I can get to the internals.

But I doubt that it is repairable since it is not even actuating when 5 V DC is applied.
However, we might get a good look inside this ultra expensive relay

@branadic
Thanks, I did send an email to Meder. Will probably get an answer after the new year.

[plesa]

You can make Xray picture at dentist

Thanks for details about 10M resistor.

[branadic]

Thanks, I did send an email to Meder. Will probably get an answer after the new year.

You can find the BT series on their website:

https://standexelectronics.com/de/produkte2/bt-bts-reed-relais/

It's a 5V Reed Relay. Is the relay located in front of the resistive voltage divider? If so, my best guess is, that they got a special 1000V version, that is not available in their catalog. Seems like the form factor doesn't match the standard reed relays.

[HighVoltage]

You can make Xray picture at dentist

Thanks for details about 10M resistor.

Well, X-Ray would not have helped.
The housing of this relay is made of ferrous / magnetic metal, 0.7 mm wall thickness
So, I took the brutal method of a saw, to take it apart.

Here are some pictures
The reed part is suited in the center core of the winding and is made out of glass
Unfortunately it did not survive my mechanical intrusion.
It seems the glass reed contact part is just inserted to the hollow core part of the coil and then everything is potted
It also seems the potting material is not of epoxy, as first thought, but of PolyUrethan material.

It has been a long time, since I have seen a modern relay in a metal housing.
I would assume it is to protect the reed relay from actuating through an outside magnetic field.
But then, where would that magnetic field come from in such an application.

It is probably the 1000 Volts rating that makes this relay so expensive.

Here are some pictures of the brutal teardown.
The last picture shows the actual contacts that used to be in a glass housing / tube

[HighVoltage]

You can find the BT series on their website:

https://standexelectronics.com/de/produkte2/bt-bts-reed-relais/

It's a 5V Reed Relay. Is the relay located in front of the resistive voltage divider? If so, my best guess is, that they got a special 1000V version, that is not available in their catalog. Seems like the form factor doesn't match the standard reed relays.

Thanks, may be they offer a free sample, if I can not buy it.
They will probably not answer before the new year.

[HighVoltage]

OK, I took some IR pictures of the R225

Instrument cold = 22°C (Room Temperature)
Instrument warm = 30°C
So, it seems the R225 is heated by about +8 degrees after 1 hour

Here are some IP pictures from my FLIR upgraded E4.

[Kleinstein]

The main reason for the high price is likely the small number of relays of this type made.

The metal can is more to protect the surrounding circuit from the stray field. It also helps to keep temperature gradients low.

The special features the relay should have the hight rated voltage, low leakage to coil and low thermal EMF.
The low thermal EMF is likely the reason for using two contacts in series.

Just for the basic function one could try something like a
MEDER LI05-1A85
though different form factor and just one contact, thus likely not low thermal EMF. 

[HighVoltage]

@Kleinstein,
Good points on the metal can and low EMF
When I get the new relay, I will take some measurements compared to a "normal" relay. This might be interesting.

@3roomlab
I did not expect those contacts to be easier available than the relay.
Thanks
Here is a picture of the reed tube contact form the website you provided.

[HighVoltage]

EDIT** and looking down the search list, i saw this
http://sg.rs-online.com/web/p/reed-relays/8674937/
http://sg.rs-online.com/web/p/reed-relays/8674927/
looks crazy similar in package. approx USD60 a pop, but MOQ is madness

Those are pretty cool relays for up to 15 kV

[dom0]

That's "COTO goodness red". The metal part looks like an external shield?!

[Kleinstein]

The external metal can also function as a magentic return path. So the relay can get more sensitive and use less power. Low power can also be important to get thermal errors down.

[HighVoltage]

I am still after the relais from Meder and may be they can supply it.
In the meantime I got the official Meder datatsheet for this relais, see attachment

MEDER Relay BT05-BV50324

[Kleinstein]

The datasheet shows the relais is a custom design for Agilent. This might also explain the rather high price.

It also shows it is rather low power and made with thermal EMF in mind. There is a thermovolateg specification, though not that low - but this could be a worst case conservative value, in the DMM enviroment this should be much lower. Also having copper alloy pins might be important to have the low thermal EMF - other reed relays might have kovar leads.

[HighVoltage]

The datasheet shows the relais is a custom design for Agilent. This might also explain the rather high price.

Yes, I talked to Meder Germany in detail about this and they confirmed it was only made for Agilent in their chech republic factory.
http://www.relmag.cz/produkty

They are "looking" now, if they still can find one of these ralais in the factory or if they can find their BOM and build one.
I am kind of amazed that I got such good quality reaction from Meder.

[HighVoltage]

After a few days, Meder came back and let me know that they can not supply the relay because of exclusivity to Keysight and they also do not have anything similar that would fit either the footprint of the other specifications. Hmm bummer. So I will get the original Keysight replacement relay.

Otherwise the unit is repaired.
I found a 10 uH inductor that was a little larger than the original but got it fitted alright with some wires and a new resistor is also installed.  A quick test shows that the meter is within calibration specs fro DC volts and OHM.

Only thing missing is the relay now and then a full calibration check.


 

[HighVoltage]

OK, a quick follow up .... Repair is successfully completed.

Parts from Keysight arrived.
There was no way around this expensive Keysight relay, as Meder would not sell me anything.
- Installed the relay
- Installed new front connector block, because the contacts did not look so nice
- Installed new low noise fan.
- Reassembled the meter.
 
Hooked up to my Geller Labs SVRT, showing: 10.00001V after warmup.
Looks like another great addition to the lab.
Thanks for all your input.

[TheSteve]

Looks great - just how many bench DMM's do you have in the lab now?