Build Details II – 74LVT16245B
Whew! Soldering all those channels must have been monotonous and boring. Maybe you used a toaster-oven reflow. We did not (too much risk to the ASIC we thought) so after installing the new channels we took a night or two off before doing the ’245. Take a break if you need, or keep going while you’re all warmed up – but the 48-TSSOP is not a job to be laughed at.
We’d like to take a moment to discuss the genealogy of the abbreviation “TSSOP”. First, there was the DIP package – Dual Inline Pin. It was a great package and could hold a big old die, but as geometry and die size shrank people wanted something smaller. The SOP (Small Outline Package, sometimes called SO or SOIC) was born. It too was a good package and could be readily soldered by hand. But high lead counts ended up in an unwieldy leadframe (the metal frame the packaged is molded onto). So a new bastard child was born, the Shrink SOIC, or SSOP. Similar proportions to the SOIC but smaller and finer pin pitch. But that device was determined to be too thick for some people (theta-JA and all) and so the THIN Shrink Small Outline Package, or TSSOP was born. It’s thin enough that you can bend it and crack the die if you want, and the lead pitch is pretty damn small. A standard DIP has 100 mil pitch, or 0.1 inch between each pin. The TSSOP has 25 mil pitch, or 4 pins for every DIP pin. GET A FINE POINT SOLDERING IRON TIP.
So here’s how we suggest soldering this beast.
Step 1 - Preparing the pads. Pre-Tin the PCB pads by painting them with flux and wandering over them with a little blob of solder on the iron tip. They should each grab a puddle of solder and look nice and uniformly covered. Now go back and remove almost all the solder using solder wick. Finally, use the iron to wipe any remaining solder out to the outside of the pads. This will create a flat surface that the IC leg can land on, which is essential – if one lumpy pad is preventing the chip from sitting flat, you can get leads that LOOK like they connect but really don’t and the debug is a bitch. The second aspect is that little reservoir of solder at the end of the PCB pad can be wiped back towards the IC to affix the leg once the chip is placed. That tiny pool is all you’ll need. If you try to bring in fresh solder, you’ll probably bridge over 3-4 of the little TSSOP leads. In that case, clean it up with solder wick and retry the wiping.
Fig 9 – Solder Management is key for the TSSOP
Step 2 – Attaching the chip. Place the TSSOP and align it. One side needs to be PERFECTLY aligned, the other side may wander and tend to drop into the spaces between the PCB pads. That’s OK as we will tack down each side seperately. Apply a little pressure to the top of the chip with your finger, and nudge/lever the body of the chip (don’t apply any pressure to the pins, they will just bend) until all legs on one side are aligned with the pads. Once that is done, use the soldering iron to tack down one corner pin. First, press the pin down with the tip of the iron – this melts the solder under the pin and the solder begins to “wet” the bottom of the pin. Now follow up by wiping from the end of the pad back up over the end of the pin. This should result in a nice sheet of solder covering the end of the pin and up over the bottom of the pin. Some more solder will also wick under the pin due to capillary action. Only do one corner pin at this time!
Step 3 – Tacking the opposing corner. Since only one corner is lightly held, the chip can be moved around a little bit to align the other side. As we said, the other pins usually fall into the gaps between the solder pads on the other side and will need to be adjusted. Stop pressing on the top of the chip for a moment and use your tweezers to adjust the chip so the other side’s pins line up with the PCB pads. We usually press the sharp tweezer points into the PCB right near the body of the chip and with a little lever action, we can precisely move the chip the fraction of a mm needed to get final alignment. With finger holding the chip in place, tack the opposing corner to freeze the chip placement.
Fig 11 – Tack the other corner once final alignment is completed
Step 4 – Solder the remaining pins, only 46 to go!. You will probably want to generously apply some liquid flux to each row of leads at this point to prevent solder bridges. Use the same technique of pressing each pin down into the pad and then wiping solder back over it. Sometimes, it is helpful to get a tiny speck of solder on the iron and rewipe over each pad-pin connection to insure it is well connected. One speck should last for 4-6 pins at least. You don’t want a drop, you just want the tip wet – no pun intended. But seriously, you can even rewet the iron by touching some other blob of solder like from one of the channels. The tiny layer of residual solder that the iron takes away is plenty for 3-4 pins.
The ’245 also need 4 bypass caps on it’s power supply. These are the pads located above and below the chip. Install 0.1uf or 0.12uF (100nF or 120nF) here, depending on what you have on hand.
Step 5 – Buzz out your board. That’s an old phrase describing the sound a multimeter makes on connectivity mode – beeping when a connection is made. It means to check the connections, of course. It’s a total pain in the ass, but we really suggest you do it to avoid blowing up a chip (or the ASIC) due to a shorted line. In order to touch the TSSOP leads, it’s best to clip onto a sewing needle or something with a very fine point. Stab the needle into the top of each of the TSSOP pins near where it enters the body of the IC. Now use the other lead of the meter to check that pin’s connectivity to the proper channel of the input circuitry. You want to check two things while buzzing out.
- Check that the pin is actually connected to the input circuit (or to the ASIC if you’re checking that side). The meter should beep (of course) when the input circuit and the TSSOP pin are touched. If not, it usually means that the bad pin is not actually soldered down to the PCB pad. That’s why you want to touch the top of the TSSOP pin – to make sure the path is good all the way from the pin, through the PCB trace, to the other circuit.
- Check that the pin is NOT shorted to any adjacent pins. This is easily done by checking the channel adjacent to the pin under test. The pin should NOT be connected to the channel before or after the present one, but it should beep nicely when the present channel is tapped.
An alternate test would be to stab the PCB pad (not the TSSOP pin anymore) and look for a GND->Pin ESD diode. By attaching the red lead of the meter to the PCB GND and stabbing each of the ’245′s PCB pads with the needle, you should see a diode with a value of about 0.5-0.6v pointing from GND to the pad. Again, you stab the pad to make sure the pin in question is well connected to the PCB pad.
You can check the ASIC side by looking at the value of the diode voltage seen. With the ’245′s diode in parallel with the ASIC’s diode the meter should read something like 0.45-0.5v. If you find one pin that is larger (such as 0.6v), then that pin is probably not connected to the ’245 and you’ll need to go back and resolder it.
It’s a good time to mention this if you have not done it already – you will need to cut the shorts between the 16 resistor pads on the back of the PCB now. Actually, you need to do this before you try doing a diode check on the ASIC side. Umm, because they would be shorting all those lines to ground and you would not see ANY diodes in that case. Here’s a pic from after we gouged out those shorts between the resistor pads on the back.
Fig 12 – Cutting the spare channel shorts on the back of the PCB
Fix any shorts by solder-wicking those pins and then resoldering. Fix any opens by repeating the press, wipe sequence and maybe another speck’s worth of solder. Buzz out again before continuing.
Continued on Next Page…
By Buzz August 18, 2010 - 4:50 am
I have created a BOM for this mod from Digikey. Does anybody want to look it over and see if there is anything that should be changed?
https://spreadsheets.google.com/ccc?key=0AnWif20cSKwEdGYtNUVJUzNvWXJkNUszSzd0SnQ5WlE&hl=en&authkey=CITJpeAM
By Matt January 2, 2013 - 7:51 pm
This response is a little late.. but the list looks great, except that the 510 ohm resistor was mistaken for 510k ohm.
By od August 29, 2010 - 4:27 am
Top site, I had not come across openschemes.com before in my searches!
Carry on the excellent work!
By MAXI888 May 7, 2011 - 8:25 am
Did someone use the mod with an actual device? Is this hack still work on it? Thx a lot for Answer.
MAX
By openschemes May 8, 2011 - 12:09 pm
Yes, definitely. We and bushing have both done the mod and it works to expand all the channels. We have heard that the chip has been changed so the recent smaller devices can no longer be expanded, though.
By Jivanildo December 28, 2012 - 7:40 am
I started years ago using a Pace MBT 250 rewrok station. Performed incredible. Cost was $1000.+The Metcal MX 500, an industry standard for many years and the new Metcal digital MX 5010. An amazing performer, truly SWEET. Cost $650. retail and cost of ownership (tips) at $30. each.The iron I prefer is the Hakko FX-951. Very compact, 70 watts, very comfortable hand piece, ultra performer, immediate thermal recovery, programmable temp. Retail $250., tips $10. each. I prefer Hakko.
By ZeroPlus LAP-C 16032 Logic Analyser « EdVoncken.NET June 27, 2011 - 1:42 am
[...] Openschemes: ZeroPlus Logic Cube – The Modification [...]
By User January 14, 2012 - 10:15 am
Did someone Know if the Hardware Hack work with the actual revision of the ZeroPlus LAP-C16032?
Thanks for the Answer
Peter
By Stu March 2, 2012 - 1:41 pm
I’m wondering, too — Is there any way to check or verify if it’s still possible to purchase a hackable device? Or have those revisions all been sold off and cleaned out?
By Zizou July 28, 2012 - 10:22 am
I have a short while ago started out your webolg, the data anyone present here possesses made it simpler for me enormously. We appreciate you your complete time period & function.
By rfasic March 27, 2012 - 8:32 am
No problems for a 32 channel extension (even with lastest asic version) but they cut 2 pins on it (asic) to prevent a RAM adressing from 32k to 128k = Fixed with 2 thin wires between PCB and Asic.
By openschemes March 27, 2012 - 5:08 pm
Congratulations! Thank you for letting us know the device is still hackable, that is great news. And nice idea splicing in the missing pins!
By vatrex April 18, 2012 - 12:17 pm
I needed to post you a little bit of word to be able to say thanks again just for the incredible guidelines you’ve documented here. It is really open-handed of you to provide extensively just what a lot of folks could possibly have sold as an e-book to earn some bucks for their own end, certainly since you could possibly have tried it if you considered necessary. The tips as well served to be a good way to know that someone else have a similar dream the same as mine to know a good deal more with reference to this issue. I believe there are a lot more pleasurable periods in the future for many who look into your blog.
By sargon April 28, 2012 - 2:51 am
having opened up my newly purchased c16032 unit made in oct 2011 , i find that there are 4 pins that are removed from the asic , cut flush , will not be an easy thing to attach back on the asic ,
the hack of doing the usbport.dll change works to allow the program to make it look like 16128 but i am wondering if because of the cut pins on the asic its working to max mem of 16128 ?
also the protocols registered will not work unless you register them with the keys for that model & serial
By openschemes April 28, 2012 - 7:13 am
Yep, not easy to solder but very similar to what Nintendo did on the Wii console for a while. To get around it, guys just ground a little of the black plastic package off and soldered to the leadframe region of the missing pins. Read about something like that here
http://www.eurasia.nu/modules.php?name=Downloads&d_op=viewdownloaddetails&lid=4284&ttitle=Soldering_guide_for_Wii_with_pins_cut
As for the protocols – yes, you are correct although you can change the serial by updating the eeprom. But if you have a lot of protocols registered to your base unit, you will not be able to use them on the modded unit.
By sargon April 29, 2012 - 8:47 pm
be good if someone who has done a current unit can advise , as even if i get the 4 pins attached again , i am still not sure the extra channels are not grounded internal to asic , have not started this adventure till i get a better tip for the Metcal
also the 7416245b and the 1.2uH are available via RS components
By Xenawise June 8, 2012 - 6:49 pm
Ordered my LAP-C(16032) from here last month:
http://microcontrollershop.com/product_info.php?products_id=3369
Just modified my LAP-C(16032) to add the 16 extra channels. The ASIC is the ZP322MC-5 with a manufacturer date of 0924 (work week 24 of 2009) and it was a success. First I cut the traced on the back of the board and measured to ground and found that they were no longer shorted so I ordered the parts to add in the extra 16 channels. The ASIC has 4 pins cut off and I still need to wire those up to make full use of the 4.5MB memory chip which is rated at 250MHz in this unit as well. The serial number would indicate this model was built in October of 2011 (1110). All I can do is verify that at this point, I was able to add on the extra 16-channels. I still have to wire up the 4 cut off pins to get full access to the memory but at this time the mod still works for me!
By openschemes June 8, 2012 - 9:46 pm
Awesome! It really is a great device, and we are thankful that you’ve found a moddable source! Thanks for posting your success, and enjoy your new 32 channel LA!!!
By CS July 10, 2012 - 10:25 am
Hello,
I can confirm that it is possible to carry out the 16 to 32 Channel modification, even with the new ZP322MC-5 ASIC! (testet with SNR 11032* and 11101*, both ASIC DC0924 !)
I also think that it is not necessary for the “32kBit/Channel” to “128kBit/channel” memory expansion to repair the four cutted PINs.
I’ve checked the layout and it seems that these four lines are the four highest address lines of the 72MB RAM.
It seems that these lines are internal “not Connectet” to the 4.5 MB Ram DIE. (128kBit/Page * 2^4 = 2048 kBit (=2MB))
thus should a repair of these four lines only be made in conjunction with a memory upgrade to 1000 or 2000 kbits /channel. (Of course – it is possible that i am wrong)
regards
-CS-
By openschemes July 10, 2012 - 2:43 pm
Wow, very exciting news!! Thanks for confirming it!
Buy em’ up people, before ZP changes it’s mind!!
By sargon June 10, 2012 - 4:36 pm
yes cutting back the silicon to solder the pins is a bit scary stuff , would like to see pics, when you done it and how you went
By openschemes June 10, 2012 - 5:00 pm
Good point. But what you actually solder to is the frame of the metal leg. Even though the leg of the chip has been chopped off, there’s at least several mm of the metal leg still within the package acting as a frame. The technique is to grind away some of the plastic package in order to expose a little bit more of the metal leg. Then, you solder to that. The silicon chip is in the center of the package, you don’t want to be grinding anywhere near that. Just near the edge where the missing leg is.
By BobP August 7, 2012 - 5:05 am
Hi All,
Just got my LAP-C16032 this morning. Opened it up and cut the resistor shorts on the bottom side of the PCB. Unfortunately the additional inputs on the ZP322MC-5 are still grounded. Mine has a date code of 0953.
I also have a LAP-A16128 which I bought thinking of modifying that one. It is missing all the tracking for the 2nd set of 16 channels but it has the older ASIC.
Apart from the date code the other marking that is different and (maybe) significant is the line on the ASIC that starts K61181.3S. My ‘old’ ASIC has IERH and this new one has 2ERH. So, if the ASIC has 2ERH on it, appears it is internally shorted on the other 16 channels.
Something else I saw on this new 16032 – the memory is upgraded speed wise. My 16032 has -250AXC memory device, so in theory could run at the full 200MHz speed of the higher spec analysers.
Am considering whether to un-solder and swap the ZP332 ASICs (have access to IR reflow machine) or whether to go with the original idea of making up a small daughter PCB for the LAP-A16128.
By openschemes August 11, 2012 - 7:38 pm
Oh no! We’re very sorry to hear that you got a bad ASIC!
Thank you for posting the identifying information for the good and bad ASIC – it’s very interesting that the memory is now at -250 speed grade. Wonder if there’s anything that this new ASIC can do that the old one can’t..?
By wivern August 13, 2012 - 1:37 pm
I have identical LAP-C16032 , ASIC K61181.3S-2ERH date 0953 and memory 250AXC.
> Am considering whether to un-solder and swap the ZP332 ASICs (have access to IR
> reflow machine) or whether to go with the original idea of making up a small
> daughter PCB for the LAP-A16128.
He can get the original ASIC as a spare part? Where and for how much?
You can develop your idea with daughter PCB?
By BobP August 17, 2012 - 11:41 am
Hi again,
Well I took the plunge and swapped the ZP322MC ASICs between the LAP-A16128 and the LAP-C16032 using an IR rework station to (carefully) remove both ASICs. PCBs were then cleaned up using flux and solder wick to get the pads flat.
Then I went round the ASIC pins getting off as much solder/ removing shorts that had occurred during the ASIC removal process.
I decided using the reork machine would be risky for replacing the ASICs – they can slip and slide so I used a fine tipped soldering iron, fine grade solder and a microscope to go round each pin individually. I tacked opposite sides first to ensure registration. Although there are a lot of pins it didn’t take too long to resolder both ASICs onto their new PCBs.Reomval, cleanup and replacement took altogether around 3 hours for both boards.
Today the final piece in the expansion jigsaw arrived – a new 74ABT16245 buffer device. Soldered that down at lunchtime and using the hacked USBPort DLL have tested each input on the expanded 32 channel 16032 device. I had failed to complete one of the cuts underneath the PCB so one channel was grounded permanently – but I found that one quick enough.
Thanks to your excellent article on how/what to do to upgrade these devices I am happy to report I have a fully functioning 16032 converted to 32128.
Just need to work out the software side now in order to re-program the EEPROM and whether to go for the memory upgrade too.
By Jack Hardball November 14, 2012 - 10:04 am
Anyone know what the part number for the upgrade memory?
By Ed March 8, 2013 - 11:03 pm
I apologize for this recent post on such an old article but I have to thank everyone for the article and all the comments. I recently (mar 2013) bought a 16032 with last four of 0102 and with a ASIC model of 0924. I was able to upgrade a LAP-C 16032 to a 322000 via eeprom and cut the traces. The other pins no longer route to ground so I need to order the parts. I was able to sand away part of the IC package to the cut pins and solder them to the traces so I will see if it all works when the memory comes in.
By openschemes March 10, 2013 - 8:26 pm
Fantastic! Would you take a picture of the completed work? Many users are uncertain about sanding down the IC to touch the cut pins.
By Ed March 11, 2013 - 4:15 am
Sure thing! I took pictures along the way haha. I’m glad it’s just the corner and not somewhere crazy like the middle of the IC, yikes! Still waiting on digikey for the RAM seems they are the only ones that carry it.
By Ed March 11, 2013 - 5:43 am
Here are the images as I mentioned earlier.
https://picasaweb.google.com/102502076918141645796/ZeroPlusLogicCube16032Mod?authuser=0&feat=directlinkI apologize if linking URLs is not allowed as I am not sure what other way to share this. On an interesting side note regarding the free 30 protocols, I came across some information on a forum and apparently the guy used his 16032 serial and verification key to register it as a 16128, as he did the eeprom change, but anciently chose 32128 and it still went through. I tried this myself using my serial number and verification key and was able to register as a 322000. I find it odd that the serial numbers are not device bound when choosing to register your model. I could see this as a good and bad thing, good for us, bad for if someone accidentally choose the wrong model in their haste. I can now use the device as a 322000, get the 30 free protocols, and have the included 10. Just my findings and testing so far. I patiently wait for the parts to come in this week and will update you on the completed (hopefully) status of it all.
A quick side question, the regarding buffer chip supply voltages, the NXP 74ABT16245BDGG I found was 4.5V – 5.5V, while the TI SN74LVT16245BDGGR has a range of 2.7V – 3.6V. Now I ordered both parts to be safe, but do you know which of the two work for this modification? and which one I should choose if both work? Thank you!
By sargon May 4, 2013 - 1:55 am
how you go ED , with the mod and the extra memory chip, can u confirm the part number u ordered ,and fitted for both memory and buffer 245
By Al May 14, 2013 - 9:18 am
Hello,
Congratulation for your successful mod !
Where bought you this lap 16032 with an ASIC model of 0924 ?
Best Regards