VOGONS


3 (+3 more) retro battle stations

Topic actions

Reply 760 of 865, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

Added your numbers.
Also added DTK PKM-0039S. An ISA/VLB main board which was a complete disaster with 486/586 processors but pod100 unlocked its potential. More details at the end of that post.

Feipoa, so not clear if refreshing old chalky soldering paste with flux is working or not ?

retro bits and bytes

Reply 761 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Hard to say right now if it is working well or not, but the consistency certainly gets refreshed and it flows out of the syringe as it should. I foolishly decided that I would experiment my pastry success on a PLCC32 socket, which if you know, is some form of ABS or poly whatever. Well, seems the hot air gut is too much for the PLCC socket, so I went to the iron. When finished, I had one solder bridge, although because of the socket, you cannot actually see it; I had to use the DMM to find it. Unfortunately, because the socket is in the way, I could not un-bridge it the hidden bridge. I had to hot air off the whole socket, which completely ruins it. So... I cannot really comment on if the paste works properly yet. It was a bad idea to test it on a PLCC socket, but that is what I was working on for another project, so I thought I'd kill two birds with one stone.

Next step is for me to dig my dead PCB box out from the back of the closet, find a dead motherboard, and try the refreshed paste on a QFP chip. I'll have to report back after that.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 762 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Attached is a photo of the new solder paste, TS391AX, applied to PLCC32 pads. I am able to apply this with relative ease. I assume this is what it should look like new. Now the old expired stuff, SMD291SNL, even with the flux paste added, I cannot apply it to these same PLCC32 pads. It doesn't want to hold onto the solder pad and more or less balls up. I am able to pull a line across all the pads with the old stuff, but nothing fine pitched like shown in that photo. I'm not sure if it is just the age or the TS391AX is generally more fluid compared to the SMD291SNL.

Attachments

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 763 of 865, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

Ok, paste looks healthy.
Going to use what you did when the one here goes dry.

Btw, you probably know, but still - i usually do one stroke with syringe across the pads grid - goes much quicker and result is the same as putting paste on each pad individually.

retro bits and bytes

Reply 764 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++
pshipkov wrote on 2021-11-30, 22:48:

Ok, paste looks healthy.
Going to use what you did when the one here goes dry.

Btw, you probably know, but still - i usually do one stroke with syringe across the pads grid - goes much quicker and result is the same as putting paste on each pad individually.

Based on your response, either I did not word my paragraph carefully enough, or you misread it. That photo immediately above is showing the NEW paste only, not the old paste with flux added. I am unable to get the old refreshed paste to do what is shown in the photo.

I only put the paste on pad by pad to demonstrate what new paste should be able to do. I cannot do that with the old paste, even with the flux added in. I didn't take a photo because it looks like a mess. Yes, normally you would put the paste on going across all the pads. As I indicated above, I am able to make a like across (perpendicular) to the pads with the old refreshed stuff, but not as thin as I'd like. and it still doesn't adhere to the surface well upon application.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 765 of 865, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

Read your responses more carefully. Makes sense.
Initially i thought this is the outcome of mixing old paste with flux. Which i thought is great, because every few months i have to buy new syringe to use few times only before it eventually goes bad and has to be replaced. Wasteful.

retro bits and bytes

Reply 766 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Let me get some comparative photos that tell the story better. File names and captions should be self-explanatory.

This time, I'm pulling a line across the pins of a QFP - basically, I am practising for the work on those Trio64V+ VLB cards where I need to solder on a giant QFP. The pitch of the QFP on this sample will be the same as the Trio64V+

Here is the bare PCB, from a SCSI controller card. I have already desoldered the QFP chip and cleaned up the pads.

1-PCB.JPG
Filename
1-PCB.JPG
File size
464.29 KiB
Views
315 views
File license
CC-BY-4.0

Here is the cleaned up QFP IC.

2-QFP.JPG
Filename
2-QFP.JPG
File size
291.23 KiB
Views
315 views
File license
CC-BY-4.0

Here, I have drawn a line across the pads using the old SMD291SNL solder paste with newly added flux stirred in. I estimate there's about 3.5 cc of paste in the syringe and that I added about 1 cc of flux paste. The light green CHIPQUIK syringe tips were the smallest I could use. I tried to use smaller, but the paste would get clogged into the nozzle. Shown here is the smallest line I could pull. Even that line, I had to do 3 takes to get it like that. You can see that the paste isn't really wanting to flow onto the PCB.

3-old_paste.JPG
Filename
3-old_paste.JPG
File size
453.63 KiB
Views
315 views
File license
CC-BY-4.0

By way of comparison, here is the new paste, TS391AX, also from ChipQuik. But to be fair, I am using the red coloured (all plastic) syringe tip. To get an idea for size difference, this is the order of tip spout diameters, from smallest to largest: red, blue, violet, light green, grey, dark green. This is not to be confused with the metal syringe tips, which would have a size ordering of (smallest-to-largest): clear, white, red, blue, violet, brown, green, black. The ID of the red plastic is about the same as the metal clear, but the metals will likely let the paste come out slower due to more back pressure (1 cm long metal shaft creates "feet of head").

4-new_paste.JPG
Filename
4-new_paste.JPG
File size
558.43 KiB
Views
315 views
File license
CC-BY-4.0

Here is the full overview. Half are with old paste, half with new paste.

5-pasted_both.JPG
Filename
5-pasted_both.JPG
File size
858.68 KiB
Views
315 views
File license
CC-BY-4.0
Last edited by feipoa on 2021-12-01, 09:37. Edited 1 time in total.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 767 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++

The view of the old and new solder paste after positioning the QFP chip.

6-old_pre-heat.JPG
Filename
6-old_pre-heat.JPG
File size
597.1 KiB
Views
315 views
File license
CC-BY-4.0
7-new_pre-heat.JPG
Filename
7-new_pre-heat.JPG
File size
248.47 KiB
Views
315 views
File license
CC-BY-4.0

This is after I ran the hot air along the pins. The old paste was a lot slower to melt compared to the new paste. Several large bridges result. This is likely due to the amount of paste put on the pads, but this is as thin as I could get it.

8-old_final_result.JPG
Filename
8-old_final_result.JPG
File size
453.92 KiB
Views
315 views
File license
CC-BY-4.0

Here's the view of the new paste after heating. There are 3 small bridges, which for the most part, correspond to where the paste was layed thicker. Obviously I need to get my line finer, which I'm hoping the metal tipped needle will help facilitate.

9-new_final_result.JPG
Filename
9-new_final_result.JPG
File size
569.3 KiB
Views
315 views
File license
CC-BY-4.0

Here's an image of the finest all plastic tip (red) and that of the finest metal tip (clear). The ID of the openings looks about the same to me, so I am guessing that the extra back pressure caused by pushing a viscous fluid through a thin pipe helps the paste eject more slowly.

Finest_syringe_tips.JPG
Filename
Finest_syringe_tips.JPG
File size
165.12 KiB
Views
315 views
File license
CC-BY-4.0

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 768 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++

In summary, I'm not sure if the SMD291SNL (old tube) product is more viscous from the onset compared to TS391AX (new tube), but I suspect even if true, it wouldn't be by much. I added a generous amount of flux to the old paste. I suspect it may only be usable for broader pitched surface mount chips. I did try adding more, testing, add more test, but after a point, the viscosity didn't continue to decrease.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 770 of 865, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

I don't know Max. Soldering this kind of chips with so many legs is a job for paste+hotgun.
It looks to me the Feipoa is 1 step away from getting the right amount of paste applied to get the desired result.

retro bits and bytes

Reply 772 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++
pentiumspeed wrote on 2021-12-01, 18:34:

More flux!

You think adding more flux to the expired tube of paste will make it flow out of the syringe like the new paste does? Add another 1 cc ?

The flux paste alone has about the consistency of the new solder paste, so it seems unlikely that I can obtain this consistency in the old paste. Does your experience indicate otherwise?

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 773 of 865, by maxtherabbit

User metadata
Rank l33t
Rank
l33t
pshipkov wrote on 2021-12-01, 16:12:

I don't know Max. Soldering this kind of chips with so many legs is a job for paste+hotgun.
It looks to me the Feipoa is 1 step away from getting the right amount of paste applied to get the desired result.

I disagree, but even if you insist on using paste and hot air, you can still clean up those bridges with some flux and a clean iron tip

Reply 775 of 865, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I agree with pshipkov. While I have done fine pitched QFP with an iron before, I feel it sufficiently beneficial to move on to a more effecient method. Optimising a new skill is also fun. I've avoided paste in the past due to the expiration dates, and from my latest tests, it does seem somewhat justified. Using paste may be the best for batch operations when you know you can use up the tube within a year. It is the pair of QFP208's from the Trio64V+/Virge VLB project, as well as wrap-under SOJ's, that convinced me it is time to move on.

When my TS391AX starts to expire, I'll try the flux trick with it as well. It may be that my old tube of SMD291SNL was just too old, or the original formula not appropriate for the added flux to benefit to the level I desired. The package did say it expires in summer 2019, which is one year from purchase. I am hoping the hype about TS391AX having a longer shelf-life is valid.

Just to clarify, adding flux paste to my old SMD291SNL did help with reducing viscosity, just it wasn't enough. Thanks to BitWrangler for this tip.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 776 of 865, by pentiumspeed

User metadata
Rank Oldbie
Rank
Oldbie

Not adding more flux to paste. Not necessary.

When you are fixing the bridges, add more flux and draw away the excess solder with your dry soldering iron by dragging the solder out. Lot of flux than too little is best, believe me. When done, clean up with 99% alcohol. Remember to blow away the smoke as you work.

The solder paste has some flux already. But quality of solder paste do matter. Should be lead solder paste. No lead solder is terrible to work with.

Cheers,

Great Northern aka Canada.

Reply 778 of 865, by BitWrangler

User metadata
Rank l33t
Rank
l33t
feipoa wrote on 2021-12-02, 22:00:

When my TS391AX starts to expire, I'll try the flux trick with it as well. It may be that my old tube of SMD291SNL was just too old, or the original formula not appropriate for the added flux to benefit to the level I desired. The package did say it expires in summer 2019, which is one year from purchase. I am hoping the hype about TS391AX having a longer shelf-life is valid.

Just to clarify, adding flux paste to my old SMD291SNL did help with reducing viscosity, just it wasn't enough. Thanks to BitWrangler for this tip.

It looked like you were still having homogeneity problems, it wasn't blending in well enough as it was looking a little blobby/stringy. Also the method you used it with was a bit "processy" in that for it to apply in a stripe and blob up nicely on the pads, it was required to have close to original properties, versus fully manual application to each pad with assistance of a toothpick or something. Could try something like a half unbent paperclip chucked in a dremel to mix very well, carefully or you'll be asking what gets it off the walls. Adding a little flux or IPA every few months as paste ages is probably the better method to keep it workable though, rather than letting it set nearly solid and trying to rehab it. If nothing else maybe you can just put it in a tiny screwtop jar and use it for tinning wires, or soldering twisted ones, just dip them then apply heat. If you want to build full PCBs, kits etc, do a larger project or small board run, then it's probably not worth the hassle not to have an in-date tube for it that behaves exactly as it should.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.

Reply 779 of 865, by Chadti99

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote on 2021-11-20, 09:08:

Do either of you have a Diamond Stealth64 DRAM-T VLB with 86c764-P chip and BIOS v2.02 and later?

Attached. Extracted with NSSI. I've not tested but it looks right, let me know!

Attachments

  • Filename
    DS764P202vlb.7z
    File size
    16.79 KiB
    Downloads
    8 downloads
    File license
    Fair use/fair dealing exception