Author Topic: How do I prototype with high-accuracy, expensive SMD ICs?  (Read 3238 times)

0 Members and 1 Guest are viewing this topic.

Offline TheUnnamedNewbieTopic starter

  • Super Contributor
  • ***
  • Posts: 1208
  • Country: 00
  • mmwave RFIC/antenna designer
How do I prototype with high-accuracy, expensive SMD ICs?
« on: May 15, 2017, 10:33:35 am »
Hi guys!

As a result of a project that we needed the part for (but we had left over) I will soon be in possesion of two Analog Devices AD5791ARUZ DACs: 20bit, low noise high accuracy DACs. These things cost about 55 Euros in 1-off quantities on Farnell.

I would like to use them (In combination with some very good reference ICs I also have) to build my own precision voltage source this summer. Usually, I would prototype this kind of stuff by doing a board, seeing if it works, and if not, start over. But as this is a 55 euro DAC, I can't just go and buy a few more if stuff breaks (because I can't warrant the cost in a personal project). I was wondering what techniques I could use to allow me to remove the part safely, without damage, in case the performance of the system is bad. One option would be to use a break outboard, and put the part on there. This should be ok in this application, since I don't think it will impact performance, but I was wondering if there were any other methods that I was unaware of that I could use for this.
The best part about magic is when it stops being magic and becomes science instead

"There was no road, but the people walked on it, and the road came to be, and the people followed it, for the road took the path of least resistance"
 

Offline sokoloff

  • Super Contributor
  • ***
  • Posts: 1799
  • Country: us
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #1 on: May 15, 2017, 11:05:10 am »
I'm not sure exactly what you're asking. SMD parts desolder fairly readily by hand with flux and braid. Get some discarded e-waste and practice on some chips in a similar package; you'll get the hang of it inside of 5 or 6 chips I think.

It's not worth doing for a $0.25 part, but for a $60 part, it's perfectly do-able. I wouldn't hesitate to go through the prototyping cycle using custom PCBs and surface mounting the chip.

If you think you'll go through a lot of spins (you're in an exploratory phase), then it might be worth making a breadboard adapter for it, but if you're just trying to get through a cycle or two (for design errors/omissions), I'd just surface mount it during development.
 

Offline jpanhalt

  • Super Contributor
  • ***
  • Posts: 3493
  • Country: us
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #2 on: May 15, 2017, 11:32:10 am »
You can get a test andburn-in socket for TSSOP devices: http://www.ebay.com/itm/FP-20-0-65-01A-SSOP20-TSSOP20-4-4Width-0-65Pitch-IC-Test-Burn-In-Socket-Enplas-/261022621476?hash=item3cc6291f24:m:m7h994NPwoZmvorMbBkdfQA

I usually get the 3M  brand (Textool), when it is available.  They look almost identical and work fine.   

John
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9891
  • Country: us
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #3 on: May 15, 2017, 04:13:30 pm »
I often just use adapters like this:

https://www.digikey.com/catalog/en/partgroup/tssop-to-dip-smt-adapters/60924

Later on, the device can be unsoldered using any of a number of techniques but I would probably use a hot air rework tool.
 

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 9526
  • Country: gb
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #4 on: May 15, 2017, 04:24:38 pm »
I would also go with one of the TSSOP adaptor boards. The only caution with a high resolution DAC (or ADC) is getting supply decoupling close up to the chip. You might want to do a little surgery to add caps to the board, or possibly straight across the dip pads if they work out correctly.
Best Regards, Chris
 

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6460
  • Country: nl
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #5 on: May 15, 2017, 04:53:31 pm »
Adapter boards will probably costing you reduction in resolution,  noisefloor and other specs.
Dont you need to seperate analog and digital ground planes for an ADC design? Any fully digital ic would be fine but an ADC ?  :-//
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12875
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #6 on: May 15, 2017, 05:05:51 pm »
Looking at its datasheet, it needs quite a lot of support circuitry - output and reference buffers, decoupling etc.   Designing a no compromises daughterboard with the DAC and support circuits might be the best way forward.
 

Offline komet

  • Supporter
  • ****
  • Posts: 155
  • Country: ch
  • Shenzhen Retroencabulator Mfg. Co.
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #7 on: May 15, 2017, 05:08:51 pm »
You could prototype with the AD5781, which has the same pinout and interface, just lacking 2 bits. It's also rather expensive though.
 

Offline Muxr

  • Super Contributor
  • ***
  • Posts: 1369
  • Country: us
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #8 on: May 15, 2017, 05:22:43 pm »
If you're dealing with components that don't have strict layout requirements meaning they can be used on a breadboard then adapter PCBs are a pretty good way to go. If the layout is very important to the operation of the device, then what I do is design a prototype module PCB for the IC in question with the bare minimum bootstrap of components.

Like if I am working on a more complex design, I will design it in sub-sections. For instance I will do power rails and voltage regulators on their separate board for prototyping, I might do the sensor block on its own board. Once I evaluate these functional modules separately.. I will then do the final design on a single board with lessons learned.

This can be useful in the future projects as well because I can then evaluate component replacements easier, or reuse these functional blocks on a new prototype.

You do end up designing more PCBs and paying more fees to your PCB house (in my case Oshpark for prototyping) but I find it saves me time and unexpected frustration. If I need to do a revision of one of the sections I can work on a different section while the new PCB arrives from the fab.
« Last Edit: May 15, 2017, 05:27:03 pm by Muxr »
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9891
  • Country: us
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #9 on: May 15, 2017, 05:23:14 pm »
Looking at its datasheet, it needs quite a lot of support circuitry - output and reference buffers, decoupling etc.   Designing a no compromises daughterboard with the DAC and support circuits might be the best way forward.

This is probably the best way to go.  Design a prototype board (daughterboard) with all the support components and arrange it with pins or terminals for prototyping.  It is pretty clear that maximum accuracy won't be achieved in a prototype setup.
 

Offline TheUnnamedNewbieTopic starter

  • Super Contributor
  • ***
  • Posts: 1208
  • Country: 00
  • mmwave RFIC/antenna designer
Re: How do I prototype with high-accuracy, expensive SMD ICs?
« Reply #10 on: May 16, 2017, 06:21:15 am »
Looking at its datasheet, it needs quite a lot of support circuitry - output and reference buffers, decoupling etc.   Designing a no compromises daughterboard with the DAC and support circuits might be the best way forward.

I think this is what I will do then. I am also thinking about giving it a shot with the cheaper version, but seeing as it's still 30 Euros... I'll have to think about it.

I'm not sure exactly what you're asking. SMD parts desolder fairly readily by hand with flux and braid. Get some discarded e-waste and practice on some chips in a similar package; you'll get the hang of it inside of 5 or 6 chips I think.

I can desolder parts easily, but I was worried that the repeated cycles would be bad for the chip. Is this not the case?
The best part about magic is when it stops being magic and becomes science instead

"There was no road, but the people walked on it, and the road came to be, and the people followed it, for the road took the path of least resistance"
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf