EEVblog Electronics Community Forum
Electronics => Metrology => Topic started by: Vacuuminded on November 07, 2017, 06:59:47 pm
-
Hi all!
Was searching online for some good low EMF shorting straps, and came up pretty empty handed. The only easily found examples were nickel plated brass, which seem to be less than ideal for below e-6. So, I have decided to have some solid copper ones milled and might have extras made if there is enough interest.
I'm thinking .032" to .050" thick, but have two basic alloys to choose from. C101, or C110 in various tempers. C101 is "oxygen free", but C110 reportedly has the highest conductivity of all copper alloys so that on the surface would seem the better choice. I don't think the level of tempering will effect conductivity or any thermal characteristics, but defer to you experts on this?? :-// At face value, it would seem the hardest temper available would be ideal, mechanically speaking.
Then there's the physical design... I'm leaning more towards "B" or "C" in the below image, what would your preference be?
Regards,
Chris
-
Shorting straps are almost always used only to connect voltage and sense lines on banana connector pairs. The sense lines carry very little current, so conductivity is not an issue. I'd go for a hard alloy. Most quality binding posts use a tellurium-copper alloy.
I would also look into getting them gold plated. When copper oxidizes the copper oxide has a VERY high thermal EMF voltage. The thermal EMF hit you take for the gold-nickle-copper transitions is usually far less than the copper oxide issue. Maintaining clean bare copper surfaces is a never-ending royal pain in the ass. You rarely see bare copper binding posts because of this.
-
Copper and a good bastards file? I'm just curious and mixing the soup again.
-
Flukes 8.5digit dmm the 8508A comes with a gold plated PCB to do this.
http://eu.flukecal.com/products/accessories/test-leads-probes-and-clips/8508a-lead (http://eu.flukecal.com/products/accessories/test-leads-probes-and-clips/8508a-lead)
Not hard to make a clone to your required dimentions, I would assume 2oz copper or more
-
Flukes 8.5digit dmm the 8508A comes with a gold plated PCB to do this.
Yes, I've been thinking about laying out a PCB to do this. The size is small enough that OSHPARK prices should be fairly competitive to the Chinese suppliers.
The shorting clips on my MV-100 reference are a bitch to get off. I don't think 2 oz copper is needed. The current in the sense inputs is very small.
-
I have never removed the shorting clips from any of my meters. I just push them out of the way.
-
... mixing the soup again.
I suspect that's a Finnish idiom that didn't quite translate*, but I also suspect we know what you mean. We'd normally say 'stirring' and it wouldn't be soup.
* Would that be a 'finished idiom'? It's OK folks, I'm here all week...
-
If you don't want fancy, a freshly cleaned length of solid core copper wire does the trick nicely. Some cores stripped out of 4mm2 electrical cable have done me nicely in the past.
-
I have never removed the shorting clips from any of my meters. I just push them out of the way.
The problem with the MV-100 is the clips don't want to push out of the way. The cutout in the clip catches on the binding post. It needs to have rounded corners.
-
It looks like a 15x35mm PCB will work fine. OSHPARK would charge you around $5 for three boards (1.6 mm FR4, 1 oz copper). I checked a couple of China fabs. I think I am going to have some built with 2 oz copper, 0.8 .. 1mm FR4, and ENIG gold finish... Looks like I can offer them for maybe $10 for 10 pieces shipped. Anybody interested in some?
I'm going to try and lay out a board tonight.
I checked the clips on my MV106. The reason they hang up on the binding posts is that they have stamped little divots on the corners of the gaps which reduces the throat of the gap to 0.255". Binding posts are 0.250" wide, so there is basically no clearance. I ground down the tips of the divots and they work well now.
-
Thanks for the input, all. I guess the consensus is that bare copper straps just aren't worth the hassle any longer, even for temporary shorts used strictly for cal/zeroing. Agreed for sense lines gold plated whatever would certainly suffice. If anybody would still want some please message me. Otherwise, looks like this project is dead. I guess advantages of a solid copper strap over a PC board or some other material must be completely mind over matter.
I thought about PC boards for this too. Gold plated would be eaay, bare copper maybe not so much as HASL or gold plating seem to be the default options nearly everywhere now.
Texaspro, so long as there is no soldermask which would make stacking them an issue, I would be interested in some at that price. If for some reason it doesn't reach fruition, I will look into doing the same.
Regards,
Chris
-
It looks like a 15x35mm PCB will work fine. OSHPARK would charge you around $5 for three boards (1.6 mm FR4, 1 oz copper). I checked a couple of China fabs. I think I am going to have some built with 2 oz copper, 0.8 .. 1mm FR4, and ENIG gold finish... Looks like I can offer them for maybe $10 for 10 pieces shipped. Anybody interested in some?
FWIW, I'd recommend hard gold for that particular application if you can find it at the right price (better wear resistance than ENIG).
-
I once did a check on what hard gold would cost on a board... it was rather eye watering. The lowest price I found was around $300 setup per run and $20+ for a hard gold edge tab. You can replace a LOT of ENIG boards for that. ENIG gold costs around $15 setup per run, regardless of board quantity.
The boards I am thinking about doing would have no solder mask. One thing is that most PCB makers want to silkscreen their part number on the board somewhere. It can be a problem finding one that wont do that.
-
I guess the consensus is that bare copper straps just aren't worth the hassle any longer, even for temporary shorts used strictly for cal/zeroing.
There's definitely a place for solid copper straps. They are hard to beat for high-end volt-nutty goodness. It would be good to have a source available at not unreasonable costs.
I think for 90%+ of applications that a PCB based design will work just fine... and it's hard to beat the cost. At a low enough price, you can afford to sacrifice them to various evil experiments.
I started looking into doing the straps a couple of weeks ago when I was working on an MV100 voltage standard and the sense straps were very difficult to remove (a little grinding fixed that). I went looking at some of my other equipment to steal a pair and found most either had captive straps that could not be fully removed, or were missing straps. I figured there would be more people in the same boat.
-
I once did a check on what hard gold would cost on a board... it was rather eye watering. The lowest price I found was around $300 setup per run and $20+ for a hard gold edge tab. You can replace a LOT of ENIG boards for that. ENIG gold costs around $15 setup per run, regardless of board quantity.
The boards I am thinking about doing would have no solder mask. One thing is that most PCB makers want to silkscreen their part number on the board somewhere. It can be a problem finding one that wont do that.
What I've seen for hard gold has been ugly as well, but I've not checked every source or for a board that small. So I certainly understand your logic.
For quick & dirty, I'd second using freshly stripped solid copper wire.
In regard to copper straps, I have copper sheet on hand, some quite thick (20oz IIRC). Thick is good for making buss bars, and thinner for trace repair.
And if you truly want/need hard gold plating, you can DIY that (i.e. nickel solution is $25 for 8oz, and hard gold starts at $99 for 2oz from Gold Plating Services (https://gold-plating.com/)).
-
Also, you can get bare copper finish on PCBs. Basically it's HASL without solder and hot air process. A lot of PCBs used to be made using the SMOBC (solder mask over bare copper) process. Now I think they usually use solder mask over nickel or tin over bare copper. A problem with plain SMOBC was pinpoint defects in the solder mask and alignment issues around pads that left bare copper exposed that could corrode.
-
Here's what they should look like. Dots are via stitching between the top and bottom layers. I'm sending them off to OSHPARK first as a test... Their CAM processor was freaking out about having just top and bottom polygons... said I had no copper layers.
-
... mixing the soup again.
I suspect that's a Finnish idiom that didn't quite translate*, but I also suspect we know what you mean. We'd normally say 'stirring' and it wouldn't be soup.
* Would that be a 'finished idiom'? It's OK folks, I'm here all week...
Hah. Indeed thank you for correction of the idiom. These similar, but not exact ones are the most hard to remember. "Mixing the pot" "Stirring the soup" .... :D To be honest I'm not sure if there is even a typical version in Finnish as there is lots of variations in use.
There is somewhere an appnote which content were partly around PCB shorting bar usage in metrology, it did also contain longish discussion of various TC effects involved. Unfortunately I can not recall the name at all.
-
... mixing the soup again.
I suspect that's a Finnish idiom that didn't quite translate*, but I also suspect we know what you mean. We'd normally say 'stirring' and it wouldn't be soup.
* Would that be a 'finished idiom'? It's OK folks, I'm here all week...
Hah. Indeed thank you for correction of the idiom. These similar, but not exact ones are the most hard to remember. "Mixing the pot" "Stirring the soup" .... :D To be honest I'm not sure if there is even a typical version in Finnish as there is lots of variations in use.
There is somewhere an appnote which content were partly around PCB shorting bar usage in metrology, it did also contain longish discussion of various TC effects involved. Unfortunately I can not recall the name at all.
"Stirring the soup" in the USA is a certain kind of activity not safe for a family-friendly forum.
-
I've ordered 6 prototypes from OSHPARK... they do offer 0.8mm / 2 oz copper, but it takes a week longer to get. I'll uses these to do some performance, fit, and finish checks and look into adding more via stitching that would not provide a fracture line across the narrow parts of the board or interfere with the banana jack contact disks.
-
These might be worth looking at too. $1 USD each plus some filing or drilling. It means I guess you would then be relying on the upper and lower surfaces and not the hole for the best contact surface.
https://meniscusaudio.com/product/gold-shorting-barspr/
-
Ready made, 3 boards and you can do 1 with 4 bananas and 2 ea with 2 bananas. https://oshpark.com/shared_projects/yOAPbUZF (https://oshpark.com/shared_projects/yOAPbUZF)
(https://www.barbouri.com/wp-content/uploads/2016/12/Shorting4PinTop01-e1483029874183.jpg)
-
I've also done a layout with longer "legs" on the "E" shape. It will let you short all 4 terminals.
I don't like shorts that use banana plugs... too many connections (some soldered, plus banana plug to jack connections are not the best), too many potentially different metals involved, and lots of thermal mass.
-
On these, at least gold flash over copper board, and I used high quality Gold bananas.
-
On these, at least gold flash over copper board, and I used high quality Gold bananas.
Mine also have heavy ENIG gold over the copper.
Soldering the bananas to the PCB introduces an unnecessary thermal EMF junction and the banana plugs add a lot of thermal mass to stabilize. I think those are the reasons why Fluke sells a PCB based 4-way shorting strap. Either way will work, but it's nice to minimize possible sources of error.
-
I have some prototypes of the 4-post straps coming from OSHPARK. They are just like the 2-post straps, except are stretched 0.75" in order to reach the second set of jacks.
If they turn out OK and I order larger quantities from China, I might do them on 1mm stock instead of 0.8 mm in order to make the longer fingers a little more rigid.
-
..., and lots of thermal mass.
Well that's one of those things that cuts both ways. A small thermal mass reaches equilibrium quickly, but that perforce implies that it all can be knocked out of equilibrium quickly. A large thermal mass can sometimes act as useful thermal lag, and function as a low pass filter on any thermal changes that do occur. Which is going to be best is going to depend on the local thermal environment and its stability on timescales similar to the thermal time constant of your short.
-
if not already known, you will most likely find some usefull information
here
https://www.eevblog.com/forum/testgear/lking-for-gold-plated-tellurium-copper-banana-plugs/msg246371/#msg246371 (https://www.eevblog.com/forum/testgear/lking-for-gold-plated-tellurium-copper-banana-plugs/msg246371/#msg246371)
https://www.eevblog.com/forum/metrology/diy-low-emf-cable-and-connectors/msg190302/#msg190302 (https://www.eevblog.com/forum/metrology/diy-low-emf-cable-and-connectors/msg190302/#msg190302)
-
Ready made, 3 boards and you can do 1 with 4 bananas and 2 ea with 2 bananas. https://oshpark.com/shared_projects/yOAPbUZF (https://oshpark.com/shared_projects/yOAPbUZF)
(https://www.barbouri.com/wp-content/uploads/2016/12/Shorting4PinTop01-e1483029874183.jpg)
Can you tell me the part number + manufacturer of those banana connectors? I've been looking for these for ages.
-
Hi, I could only get with M3 stud from RS Electronics:
Multi Contact, 4mm Banana Plug, Gold Plated, 50A
Part details
Manufacturer Multi Contact
Mfr. Part No. 22.1055
https://media.rs-online.com/t_large/F2622028-01.jpg (https://media.rs-online.com/t_large/F2622028-01.jpg)
-
Just to point out that Multi-Contact have quite a range of 4mm studs, well worth a look at the catalogue if it's something you ever have a need for.
-
Hi, I could only get with M3 stud from RS Electronics:
Multi Contact, 4mm Banana Plug, Gold Plated, 50A
Part details
Manufacturer Multi Contact
Mfr. Part No. 22.1055
https://media.rs-online.com/t_large/F2622028-01.jpg (https://media.rs-online.com/t_large/F2622028-01.jpg)
-
:) They cost around $70 US Dollars for the Keithley parts. The Pomona Shorting bar costs around $3.30 US Dollars. :)
-
:) They cost around $70 US Dollars
Ouch! My 4-wire ones should run $2-$3 each. I'm shooting for $10 for 10 of the two wire ones. I might do a combo pack of 6 x 2-wire and 4 x 4 wire ones for $10.
-
Hi, I could only get with M3 stud from RS Electronics:
Multi Contact, 4mm Banana Plug, Gold Plated, 50A
Part details
Manufacturer Multi Contact
Mfr. Part No. 22.1055
https://media.rs-online.com/t_large/F2622028-01.jpg (https://media.rs-online.com/t_large/F2622028-01.jpg)
Thanks. But now I've found the Multi-contact 22.1054 which is gold plated, and metric, and 1/3 the price of the keystone.
Anyway, If you are still looking for banana shorting, I've used the Pomona MDP-S before. Which is only shorting 2 terminals, but it is stackable, so 3 of them should short your meter, if the distance is standard.
-
Hi, I could only get with M3 stud from RS Electronics:
Multi Contact, 4mm Banana Plug, Gold Plated, 50A
Part details
Manufacturer Multi Contact
Mfr. Part No. 22.1055
https://media.rs-online.com/t_large/F2622028-01.jpg (https://media.rs-online.com/t_large/F2622028-01.jpg)
Thanks. But now I've found the Multi-contact 22.1054 which is gold plated, and metric, and 1/3 the price of the keystone.
Anyway, If you are still looking for banana shorting, I've used the Pomona MDP-S before. Which is only shorting 2 terminals, but it is stackable, so 3 of them should short your meter, if the distance is standard.
Look carefully at the datasheet for the Multi-contact 22.1054 ... In Newark's system the contact material is "Copper" but the datasheet says "CuZn" which is brass. That's probably not what you want.
-
Hi, I could only get with M3 stud from RS Electronics:
Multi Contact, 4mm Banana Plug, Gold Plated, 50A
Part details
Manufacturer Multi Contact
Mfr. Part No. 22.1055
https://media.rs-online.com/t_large/F2622028-01.jpg (https://media.rs-online.com/t_large/F2622028-01.jpg)
Thanks. But now I've found the Multi-contact 22.1054 which is gold plated, and metric, and 1/3 the price of the keystone.
Anyway, If you are still looking for banana shorting, I've used the Pomona MDP-S before. Which is only shorting 2 terminals, but it is stackable, so 3 of them should short your meter, if the distance is standard.
Look carefully at the datasheet for the Multi-contact 22.1054 ... In Newark's system the contact material is "Copper" but the datasheet says "CuZn" which is brass. That's probably not what you want.
"Lantern-style machined from brass and gold plate. The lantern spring is drawn from copper alloy sheet to provide constant contact pressure and a low resistance."
-
OSHPARK has the prototype straps off to the board fab. They should be back around Dec 8.
-
OSHPARK has the prototype straps back from the fab. They should be here in a few days. If they check out, I'll send the boards off to China for some quantities. I did both two post and four post straps.
-
Got a little stash of these arrived today. 1mm Gold flashed Brass based on slightly magnetic and double spacing 38mm or 1 1/2"
-
I'd be a little concerned about thermal EMF issues with brass / steel.
OSHPARK has shipped mine. They should have been here today but did not arrive.
-
Steel certainly. Brass a little bit. I do have a little bit of Silver Bar left from making High powered batteries but my 'source' of that is no more.
-
I got the boards back from OSHPARK... they did not have any copper on them! Grrrrr...
-
I got the boards back from OSHPARK... they did not have any copper on them! Grrrrr...
It turns out the problem was the polygon boundary line was set to 0 width which caused OSHPARKs .brd processor to time out when generating the Gerbers. Setting the width to 0.001" fixed it.
The HP3458A has a slightly wider horizontal banana spacing than the vertical spacing and the 4-post strap was rather tight when inserted from the top of the banana jacks. It fit fine when inserted from the sides. I tweaked the spacing. I also added a row of via stitching down the middle of the strap.
-
Nice, I did not recognize this topic when I finally designed my 4-wire short (https://www.eevblog.com/forum/oshw/4-wire-short/):
(https://github.com/EleDes/4-wire-short/blob/master/19mmx19mm/images/4-wire%20short%20square%204mm%20top%20front%20assembled.jpg?raw=true)
I thought about modifying this for the larger space 4-wire terminals (HP 3457A, 3458A, ...), but decided to not do this simple modification, because a simple PCB screwed down should give lower emf and lowers the price.
Is there already a PCB-design for this?
My design idea (easy to modify for 19x19mm):
-
I just got in the new shorting strap PCBs from OSHPARK. There is a 4 wire and two wire version. I'm not going to be selling the boards directly but you can get them here for cheap:
Four wire:
https://oshpark.com/shared_projects/j4Htt3KM
Two wire:
https://oshpark.com/shared_projects/QBeVyK9X
You should use the 0.8 mm board thickness / 2oz copper option. If you need more than three boards it can be cheaper to download the Eagle .brd file and send it to dirtypcbs.com... specify ENIG gold finish.
-
bump