Author Topic: Yet another resistance decade box  (Read 101206 times)

0 Members and 1 Guest are viewing this topic.

Online Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: Yet another resistance decade box
« Reply #325 on: August 09, 2016, 11:34:06 am »
Hi Alex,

Thanks for sharing your measurements. I was also planning on arranging my resistors as you described toward the end of your post (lowest error for 1, next-lowest for 2, etc.).

That is not the best approach, as you need to take into account the sign of the errors. For example, you have three resistors, 10.0015K, 10.002K and 9.997K, with the errors of +1.5, +2 and -3 Ohm. The most accurate arrangement will be 10.0015K, 9.997K and 10.002K, not in the order of the error magnitude! So each next resistor should be chosen for the smallest possible total resistance error at that position.

Cheers,

Alex
 

Online Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: Yet another resistance decade box
« Reply #326 on: August 09, 2016, 02:14:22 pm »
A little update. After measuring the accuracy of the decades, I've decided to build up the better one up to full 6 digits and after finishing it, just for fun I set it as close as possible to the measured 1000000 ohm value and left HP3458A logging data at NPLC100. Here is the result for 2.5 hours and some photos. As you may see, the difference between the set and the actual value here is 23 ohm, or 23ppm from 1M.

Cheers

Alex

« Last Edit: August 09, 2016, 02:23:07 pm by Alex Nikitin »
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 9057
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: Yet another resistance decade box
« Reply #327 on: August 10, 2016, 12:50:04 am »
That is not the best approach, as you need to take into account the sign of the errors.

Yes, certainly. The way I wrote it wasn't all that clear.
TEA is the way. | TEA Time channel
 

Offline ivaylo

  • Frequent Contributor
  • **
  • Posts: 661
  • Country: us
Re: Yet another resistance decade box
« Reply #328 on: August 12, 2016, 07:08:57 am »
@Fortran, got mine, thank you so much...
 

Offline Fortran

  • Regular Contributor
  • *
  • Posts: 206
  • Country: fi
Re: Yet another resistance decade box
« Reply #329 on: August 12, 2016, 07:17:23 am »
Great :)

There's still lots of 1% kits left if anyone wants.
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 9057
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: Yet another resistance decade box
« Reply #330 on: August 12, 2016, 06:26:22 pm »
There's still lots of 1% kits left if anyone wants.

I'm telling friends about it and to post here if they want to order.
TEA is the way. | TEA Time channel
 

Offline Back2Volts

  • Supporter
  • ****
  • Posts: 495
  • Country: us
Re: Yet another resistance decade box
« Reply #331 on: August 13, 2016, 02:28:56 am »
A little update. After measuring the accuracy of the decades, I've decided to build up the better one up to full 6 digits and after finishing it, just for fun I set it as close as possible to the measured 1000000 ohm value and left HP3458A logging data at NPLC100. Here is the result for 2.5 hours and some photos. As you may see, the difference between the set and the actual value here is 23 ohm, or 23ppm from 1M.

Cheers

Alex

I would send it back!    :-DD
 

Offline Back2Volts

  • Supporter
  • ****
  • Posts: 495
  • Country: us
Re: Yet another resistance decade box
« Reply #332 on: August 13, 2016, 02:38:28 am »
Fortran, the kits came in the mail today.    Thanks!

(waiting for some little boxes from China)
 

Offline kripton2035

  • Super Contributor
  • ***
  • Posts: 2572
  • Country: fr
    • kripton2035 schematics repository
Re: Yet another resistance decade box
« Reply #333 on: August 13, 2016, 01:11:29 pm »
suggestion for PeterZ and Fortran :
do the same operation for BCD wheels and capacitor boxes
like explained here : http://kripton2035.free.fr/LabTools/lbt-capacitordec.html

thanks !
 

Offline ez24

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
  • L.D.A.
Re: Yet another resistance decade box
« Reply #334 on: August 13, 2016, 07:38:32 pm »
suggestion for PeterZ and Fortran :
do the same operation for BCD wheels and capacitor boxes
like explained here : http://kripton2035.free.fr/LabTools/lbt-capacitordec.html

thanks !
Second this
YouTube and Website Electronic Resources ------>  https://www.eevblog.com/forum/other-blog-specific/a/msg1341166/#msg1341166
 

Offline smgvbest

  • Supporter
  • ****
  • Posts: 630
  • Country: us
    • Kilbourne Astronomics
Re: Yet another resistance decade box
« Reply #335 on: August 15, 2016, 11:35:53 pm »
Finally got the 3D Printer fixed and printed up my enclosure.  it needs some tweaks of course.
I printed a sample with a 30% infill and the screw holes did not work to good so final will need to be printed with a greater infill percentage.

Here's photos of the printed case and assembled.

While I like the front and back plate idea,  I like the print quality better when its printed whole. 
once the tweaks are done I'll release the STL files for it. 




Sandra
(Yes, I am a Woman :p )
 

Offline ez24

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
  • L.D.A.
Re: Yet another resistance decade box
« Reply #336 on: August 16, 2016, 12:07:35 am »
Finally got the 3D Printer fixed and printed up my enclosure.
Looks nice, wish I had one (hint hint)

Edit:  how can you refuse someone with such a pretty forum picture?
« Last Edit: August 16, 2016, 03:46:04 am by ez24 »
YouTube and Website Electronic Resources ------>  https://www.eevblog.com/forum/other-blog-specific/a/msg1341166/#msg1341166
 

Offline Back2Volts

  • Supporter
  • ****
  • Posts: 495
  • Country: us
Re: Yet another resistance decade box
« Reply #337 on: August 16, 2016, 12:57:39 am »
Finally got the 3D Printer fixed and printed up my enclosure.  it needs some tweaks of course.
Neat !   Very neat!
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 9057
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: Yet another resistance decade box
« Reply #338 on: August 16, 2016, 03:43:23 am »
That case looks great. I assumed that you were printing the base in one piece, just selecting the desired front/rear panel before printing.
TEA is the way. | TEA Time channel
 

Offline smgvbest

  • Supporter
  • ****
  • Posts: 630
  • Country: us
    • Kilbourne Astronomics
Re: Yet another resistance decade box
« Reply #339 on: August 16, 2016, 11:14:50 am »
Thank you for the comments

The front/back panels are actual snap/slide in panels that are printed separately.  There are guides in the bottom part and up the sides. then the top part has groves to lock it in place.  it's actually very strong this way and no give on the panels when you push on them.   the walls are 1.5mm thick for the front and back panels.    When I printed them i did so with them laying flat mainly so the embossing would print nicer.  I may try a print on edge and see how that comes out.

on the other side of this.  I got it fully assembled and my 10 decade box has a 1.5ohm total contact resistance.  that's 150mOhm per decade averaged.


Sandra
(Yes, I am a Woman :p )
 

Offline julian1

  • Frequent Contributor
  • **
  • Posts: 721
  • Country: au
Re: Yet another resistance decade box
« Reply #340 on: August 20, 2016, 10:59:27 pm »
Thanks Fortran, boards arrived well, and look really nice!
 
 

Offline Blastcap

  • Regular Contributor
  • *
  • Posts: 52
  • Country: de
Re: Yet another resistance decade box
« Reply #341 on: August 23, 2016, 07:45:20 am »
Thanks a lot Fortran!
As everyone else is saying, the packaging was perfect and professional, shipping lightning fast.
The boards are really nice, with a sexy blue stop mask.

I had to file a little of edges of the board to make them fit, and the make the chamfering on the corners a little bigger, but that was probably because my thumbwheels had some molding errors.

you can make a decent little decade resistance with those boards, and for our German viewers, here are the parts used to make it, and what it costs.

0,1% Kit from Frotran, 20€
https://www.eevblog.com/forum/projects/yet-another-resistance-decade-box/msg978465/#msg978465
Housing 1,76€
https://www.conrad.de/de/1423193-534346.html
Bannana plug/Binding post 2x 1,72€ (= 3,44€)
https://www.conrad.de/de/sicherungs-steckverbinder-schwarz-10-a-sci-1-st-732314.html
Thumbwheels 4,29€
http://www.aliexpress.com/item/10pcs-Black-16mm-x-6mm-0-9-Digits-BCD-Code-Pushwheel-Thumbwheel-Switches-KM2/32698134731.html


Total material cost for a 6-digit 0,1% Decade resistace:
29,49€
 :wtf:
At that price everyone should have one... or three.
I know i'm going to build a bunch more.


edit:
Sorry for the messy workspace, i know, i know.
A 1% box would end up costing only 15,49€!
« Last Edit: August 23, 2016, 07:56:44 am by Blastcap »
 

Offline Fortran

  • Regular Contributor
  • *
  • Posts: 206
  • Country: fi
Re: Yet another resistance decade box
« Reply #342 on: August 23, 2016, 10:38:11 am »
Glad you like it :)
A decent decade it is indeed.  Not really professional, but for the price it's a lot of bang per buck.
My first decade was made of similar switches but with the original boards, and 5% resistors.
I'd be better off with a potentiometer and a multimeter :)


I have been toying with the idea of bodging a trimmer across the first resistor on the first boards to improve performance a tad more.
By reducing the contact resistance (of ONE board) (times TWO) on the 1R 10R and 100R setting, you would get an optimal performance at 111ohm.
Simulations show that about 50xR trimmers put you somewhat right in the 50% zone they are delivered in.
Over 1K, half an ohm of contact resistance doesn't matter (to me at least).
Any opinions on that?

Also.. since I still have a lot (lots and lots) of boards left and the 0.1% kits seems to be all the rage these days, I'm making another run on those.
I need 10 kits booked to justify an order, so be sure to notify me if you want one.
1% kits are still available.
 

Offline Blastcap

  • Regular Contributor
  • *
  • Posts: 52
  • Country: de
Re: Yet another resistance decade box
« Reply #343 on: August 23, 2016, 10:58:31 am »
How bout just having the 1st resistor on the 1 ohm range be a 0,5ohm? That would offset the contact resistance i've measured in mine to a reasonable precision.

But i like your idea of a trimmer across the first.
Stay tuned i might just order a buttload (that would be 20pcs) boards, and 1%'ers to play with, and do just that.
 

Offline Fortran

  • Regular Contributor
  • *
  • Posts: 206
  • Country: fi
Re: Yet another resistance decade box
« Reply #344 on: August 23, 2016, 11:15:14 am »
Well.. if you want 47 ohm then?
or 150?

111 seems to be the best compromise between all ranges but I'm open to suggestions :)
 

Offline ez24

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: us
  • L.D.A.
Re: Yet another resistance decade box
« Reply #345 on: August 23, 2016, 09:59:12 pm »
Thanks a lot Fortran!
Housing 1,76€
https://www.conrad.de/de/1423193-534346.html
Bannana plug/Binding post 2x 1,72€ (= 3,44€)
https://www.conrad.de/de/sicherungs-steckverbinder-schwarz-10-a-sci-1-st-732314.html

I have a couple of switches looking for a home.  Are there any Aliexpress or US Amazon sources for a box like the above (I cannot read German).  If it has been mentioned before and you know where, please list again.  Currently there are over 300 posts on this subject.

What I suggest for Fortran is come up with a list of sources for boxes, switches, plugs, whatever and copy and paste it into EVERY post that he makes under this subject.  That way every time there will be a source of parts.  I think it would be hard for someone new to this to figure out what is going on (because of the large number of posts).  It took me awhile to figure this out (just the resistors and switches, not box)

Or better yet start a new more focused topic.  I think this topic is a poor way to sell these, and since they are "good", this is a pity.

Something like "DIY Resistance Decade Box using SMD resistors - List of sources - $xx to $xx"

Then write a summary and list sources for everything, this topic is getting too long.


YouTube and Website Electronic Resources ------>  https://www.eevblog.com/forum/other-blog-specific/a/msg1341166/#msg1341166
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 9057
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: Yet another resistance decade box
« Reply #346 on: August 23, 2016, 11:12:51 pm »
A thread like that could also be a good place for folks to show off their finished versions/variations.
TEA is the way. | TEA Time channel
 

Offline Blastcap

  • Regular Contributor
  • *
  • Posts: 52
  • Country: de
Re: Yet another resistance decade box
« Reply #347 on: August 24, 2016, 07:22:03 am »
Well.. if you want 47 ohm then?
or 150?

111 seems to be the best compromise between all ranges but I'm open to suggestions :)

I might have expressed myself badly.
Say the first decade is assembled with 8 normal 1 ohm resistors, and the first one in the range is trimmed to be 1 ohm minus contact resistance, so for a total contact resistance of 0,5 ohm the first resistor to contact in the 1 ohm row would be a 0,5 ohm, thus when set to 1 ohm it would just add enough to the contact resistance to reach a total of 1 ohm, and after that its all 1 ohm resistors that are added.

This way 47 wouldn't be a problem (40+0,5+(6*1)), and 111 would end up being 111 even thou you only switch in 110,5 ohm.
150 on the other hand wouldn't have the contact resistance trimmed out, as it doesn't use the 1 ohm range, so it would be 150,5 ohm in reality.
So that is a lack of course.
 

Offline Fortran

  • Regular Contributor
  • *
  • Posts: 206
  • Country: fi
Re: Yet another resistance decade box
« Reply #348 on: August 24, 2016, 08:35:59 am »
Yes that's why I figured one third of the total contact resistance on the first three digits would be good.
I hardly ever use the 1R range. But 10 and 100 are more frequently used.
So if total resistance is 0.5 ohm, setting 1000 would give 1000.5 ohm. This is fine since 0.1% of 1K is 1R.

Setting 1 would give you 1.33 ohm.
Setting 150 would give you 150.17 ohm.
Setting 147 would give you 147.0 ohm.

Something like "DIY Resistance Decade Box using SMD resistors - List of sources - $xx to $xx"
Not a bad idea. That way I could keep the first post updated with the most relevant information.
Should I post in the for sale section?
 

Online Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1150
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: Yet another resistance decade box
« Reply #349 on: August 24, 2016, 01:07:13 pm »
Here is one of my finished decades.

Cheers

Alex

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf