KX Reference

[CalMachine]

Rather than further digging the hole, that is the 80-page LTZ1000 thread, I've decided to start a new thread for any members to showcase their KX references designed by TiN!

This weekend I was able to assemble everything, but 1 component, on 1 of my KX references.  Somehow I overlooked the 22 nF cap when placing my order from Digikey... so I will have to wait a few days for that to arrive until it is finished and I can start monitoring.  I am still really new to SMD soldering, so just assembling this board took me several hours.  I didn't leave my iron on for more than 2 or 3 seconds before removing heat and applied cooling.  I used tweezers/forceps for extra heat dissipation where applicable and the solder I used was SN62/PB36/AG2

I purchased Vishay 'black magic' resistors for both my my KX boards, but I decided to get some PWW resistors from Edwin for my 2nd KX reference.

Major components ready to go!


Clean board ready to be assembled.


Components from Digikey... sadly there is no 22 nF caps in there 
[/url]

Small surface mount resistors were put on first.


First AZ OP and caps put into place


Second AZ OP and MAX temp sensor put in place
[/url]

Getting the magical resistors in place!  Also, the underside MAX Temp sensor was put in place prior to this.  I had forgotten to take a picture at that step.


Heater resistors and Zener current resistors are in.  There was no way these metal enclosures were going to fit side by side, so I had to improvise slightly.  I am not certain, but I feel like my solution was a pretty good one.  I did not have to bend leads much.  Once one resistor was in the holes to the approximate depth I was looking for, I grabbed the leads between the PCB and resistor body for support.  All stress of bending was put into the leads, and leads alone.  I've seen other members wrap some sort of foil or something around the heater pair for increased thermal coupling.  I will most likely end up doing something similar to that once I figure out what I want to do.


LTZ1000 tac'd down!  Man... I've never been more nervous or scared while soldering, in my life!  Nothing like getting your blood pumping, than soldering on a $50 component for the first time.


The underside after soldering.


Tada!!!!!










Here is a little glass dome that I was wanting to help keep the LTZ1000 at thermal equilibrium as well as isolated from any air movement.  Although,  I've read that TOO much isolation can be bad.  I would like to hear opinions on a small little glass dome, such as this one.





I would love to hear any questions/comments/concerns regarding this board.  I am still a very novice VoltNut, so any advice would be greatly appreciated

[TiN]

Glass dome is cool, where you got it? Too much insulation might be case for LM399, but for LTZ I wouldn't worry, there is enough thermal conductivity on FR4 board anyway.
I sense some hate rays were sent when assembling with those open guard traces, aye?

Check those film caps for shorts, it's VERY easy to destroy them by soldering.

[CalMachine]

Glass dome is cool, where you got it? Too much insulation might be case for LM399, but for LTZ I wouldn't worry, there is enough thermal conductivity on FR4 board anyway.
I sense some hate rays were sent when assembling with those open guard traces, aye?

Check those film caps for shorts, it's VERY easy to destroy them by soldering.

I found the glass domes on flebay, actually.   I just made sure the internal dimensions were sufficient to house the LTZ1000 can.  They seem like they are almost a perfect fit!  I am still uncertain how I want to stick the dome to the board....  adhesive? little foam strip with small tie-down over the board?
http://www.ebay.com/itm/Hollow-Glass-Dome-for-Jewelry-10-5x19-5mm-Half-Dome-4-Pieces-/391731757716?

On multiple occasions I caught myself shouting "What the fuck?!"    But in the end, I feel like it came out very well.   

I will definitely check the caps.  Those were probably one of the more difficult components to solder.  Somehow I managed to order the 1210 package for the 0.1 µF caps instead of the 1206...  so, that was fun!

[ManateeMafia]

Welcome to the KX LTZ club. Now you will have to make room somewhere and dedicate a meter to capturing data. xDevs will be wanting some ppm data soon.

I will be watching your progress. Edwin has an order from me so my project will be taking my time and money.

[Kleinstein]

The glass dome looks cool, but the whole thing should be inside a metal case anyway. So not much to see. Some (those made in larger quantities) glass ware is surprisingly cheap at the chemists supply.

When making a second version with PWW resistors, keep in mind the larger form factor - so it should use a different board layout, as bending the leads to much is not a good idea.

[Dr. Frank]

The glass dome gives the LTZ1000 a nice look, as if it would be the kilogram prototype, Le Grand K.

Another thermal and air shield on the solder side would be much more important.

Frank

[CalMachine]

Welcome to the KX LTZ club. Now you will have to make room somewhere and dedicate a meter to capturing data. xDevs will be wanting some ppm data soon.

I will be watching your progress. Edwin has an order from me so my project will be taking my time and money.

I look forward to seeing how the reference performs     Right now, I've only got a Datron 1082 in my home lab for logging.  It's still a pretty phenomenal 7.5 digit meter with a DCV spec of ~10ppm/yr on the 10V range, but no 3458... 

I have yet to place my order with Edwin, I need a little bit of time to replenish funds. 

The glass dome gives the LTZ1000 a nice look, as if it would be the kilogram prototype, Le Grand K.

Another thermal and air shield on the solder side would be much more important.

Frank

I agree, I plan on having it reside in a die-cast aluminum enclosure for its life.  It might not be much to look at while it's in there, but outside it does look rather awesome! 

I've noticed other KX references with PWW resistors have some lead bending, and placement, that I am wanting to try to do... so with that said, I might rethink my KX referenced with PWW resistors.  I might just stick the Vishays in there and save the PWW resistors for my own design, instead of having 2 KX references with different resistor types.

The glass dome gives the LTZ1000 a nice look, as if it would be the kilogram prototype, Le Grand K.

Another thermal and air shield on the solder side would be much more important.

Frank

Noted!  I can another dome on the bottm.  Do you think applying a small ring of some sort of foam, for the domes to sit on, would be a bad idea?

[ManateeMafia]

With that glass dome it reminds me of Gallifrey. If you are a Dr. Who fan then even better. Of course that would be more of a reference for a time-nut.

[mimmus78]

Isn't the dome too big and too far away to prevent air drafts?

My guts tell me the dome must be near the ltz1000 ... the less air there is, the less gradient there is, the less drafts there be.

But I maybe completely wrong there.

Inviato dal mio Nexus 6P utilizzando Tapatalk

[CalMachine]

Isn't the dome too big and too far away to prevent air drafts?

My guts tell me the dome must be near the ltz1000 ... the less air there is, the less gradient there is, the less drafts there be.

But I maybe completely wrong there.

Inviato dal mio Nexus 6P utilizzando Tapatalk

I would think the dome would cause the air inside to be stagnant, right?  And just sit there, slowly reaching equilibrium with the air outside of it?  I could very well be wrong though.

With that glass dome it reminds me of Gallifrey. If you are a Dr. Who fan then even better. Of course that would be more of a reference for a time-nut.

I tried but couldn't really get into it.  The idea came to me while reading through the LTZ1000 thread.

[Vgkid]

I would think the dome would cause the air inside to be stagnant, right?  And just sit there, slowly reaching equilibrium with the air outside of it?  I could very well be wrong though.

I was thinking the same. It is only about 2cm^3, so not much volume. I would also assume that the ltz package would take up most of that.
---I never liked Dr. Who.

[IconicPCB]

Doctor Who is OK when viewed purely as a bit of sci fi.
But then so is Startrek...
HOWEVER... neither can be ignored on basis of their contribution to social and political commentry.

[mimmus78]

>> I would think the dome would cause the air inside to be stagnant, right?
>> And just sit there, slowly reaching equilibrium with the air outside of it?

Well maybe this is good experiment to do with the low noise amplifier and
check if there is differences ... actually I'm a lot busy and than I don't have
the dome.

[Kleinstein]

The question with the size of the air space above the reference is, if this is large enough to develop a significant  convection flow, especially if it even turns turbulent. So 2 cm² might be just acceptable, but a little less might be better.

With the wire would resistors, I would definitely use a different layout made for the larger form factor: Bending the wires so much to fit them to the KX board can ruin the stability and if mounted vertically, it can cause thermal EMF problems. The other point to change in that step would be to go back to the more suitable LT1013 instead of 2 single OPs. AZ OPs here can cause more trouble than good. Similar to the drift of the resistors, offset drift of the OPs is attenuated by a factor of about 200.

[CalMachine]

The question with the size of the air space above the reference is, if this is large enough to develop a significant  convection flow, especially if it even turns turbulent. So 2 cm² might be just acceptable, but a little less might be better.

With the wire would resistors, I would definitely use a different layout made for the larger form factor: Bending the wires so much to fit them to the KX board can ruin the stability and if mounted vertically, it can cause thermal EMF problems. The other point to change in that step would be to go back to the more suitable LT1013 instead of 2 single OPs. AZ OPs here can cause more trouble than good. Similar to the drift of the resistors, offset drift of the OPs is attenuated by a factor of about 200.

I'll see if I can get some smaller glass domes for relatively cheap.

And I agree, with my own design I will be doing, I plan on using the LT1013A.  I'll be sure to design enough room for bulky WW resistors when I get to laying a board out.

[mimmus78]

I used insulating foam (the one used in walls). It's more dense than standard padding foam and very easy to "machine/cut" with a sharp knife. I think it's also not so much flameable ...

Inviato dal mio Nexus 6P utilizzando Tapatalk

[Zucca]

Next Chapter: Vacuum Chamber for the ltz package 

Maybe a PVC Block with an air valve screwed on it and seal with a O Ring?

some inspirations...just writing without thinking too much... in this boring work day..

http://heater.belilove.com/article_192_Circuit-Board-Mounted.cfm

[kj7e]

I keep thinking, why not just place the entire board in an enclosed, thermally stable temperature regulated box?  So all of the components are  at a constant temp.

[Conrad Hoffman]

I can't think of anything good about putting the board in a vacuum chamber. Better to fill the chamber with dry nitrogen or argon, or immerse everything in an oil bath.

[tggzzz]

I keep thinking, why not just place the entire board in an enclosed, thermally stable temperature regulated box?  So all of the components are  at a constant temp.

Some do just that, e.g. my vintage Trancell-II. https://www.eevblog.com/forum/metrology/standard-reference-labs-trancells-any-information-available/

That has a zener in a temperature-controlled box, plus a mains supply in a different box, plus gel-cells so it isn't reliant on the mains supply - all thermally insulated in an instrument case the size of an Agilent DMM.

[CalMachine]

I keep thinking, why not just place the entire board in an enclosed, thermally stable temperature regulated box?  So all of the components are  at a constant temp.

This is what I will be doing in my own personal design. 

[Conrad Hoffman]

CalM, a friendly request- I love pictures of boards, parts and teardowns, but if you could downsize these slightly it would greatly help my steam-powered antique computer and economy 'net connection. All my money goes into test equipment and health care! Here's a 4 meg image downsized to under 350 k. Quality loss is, I think, minimal.

[CalMachine]

CalM, a friendly request- I love pictures of boards, parts and teardowns, but if you could downsize these slightly it would greatly help my steam-powered antique computer and economy 'net connection. All my money goes into test equipment and health care! Here's a 4 meg image downsized to under 350 k. Quality loss is, I think, minimal.

Yes!  I will see what I can do here, soon.   I apologize for the massive pictures

[babysitter]

It needs more JPEG!

[PedroDaGr8]

Here you go.