TDS754D - solder advice, and some whining

[TAMHAN]

Hello Folks,
its time to get back to my TDS754D - if just to clean up the lab, as the DPO takes one full work bench. Due to the presence of an error targeted specifically at U416, I removed the IC. Turned out to be somewhat simple: first, I cut one row of pins loose, and then I broke the IC out and cleaned the pads a bit.

In the next step, I figured out to try my replacement parts...only to see that I had bought SOJ400 instead of SOJ300 (bummer). But with the package already open, I cant return them to Mouser either way.    , I am an Idiot. Need to figure out what to do with these SRAMs now, especially as they came with a large warning about moisture when exposed.

But on to the main problem. All I have is an ERSA i-con nano. I soldered some normal SMD stuff before, but never one of these SOIC ICs which have the pins bent under the main IC body.

Anyone, any ideas for me?

All the best
Tam

[TAMHAN]

Hello,
it looks like the Forum ate my second image showing the actual IC. So here it is.

[Jwalling]

Hello Folks,
its time to get back to my TDS754D - if just to clean up the lab, as the DPO takes one full work bench. Due to the presence of an error targeted specifically at U416, I removed the IC. Turned out to be somewhat simple: first, I cut one row of pins loose, and then I broke the IC out and cleaned the pads a bit.

In the next step, I figured out to try my replacement parts...only to see that I had bought SOJ400 instead of SOJ300 (bummer).

Holy crap!  I thought I was the only one to make the same mistake repairing TDS700 scopes. I still have a full tube of the SOJ400 chips collecting dust.

But with the package already open, I cant return them to Mouser either way.    , I am an Idiot. Need to figure out what to do with these SRAMs now, especially as they came with a large warning about moisture when exposed.

But on to the main problem. All I have is an ERSA i-con nano. I soldered some normal SMD stuff before, but never one of these SOIC ICs which have the pins bent under the main IC body.

Anyone, any ideas for me?

All the best
Tam

You need a very fine point soldering iron, and very thin solder. I tin one pad then set the chip on it and solder it. After making sure that the alignment is good, then just solder the rest.

Jay

[eKretz]

Same here with the soldering method - I find it extremely helpful to use some extra paste flux as well. It helps a lot with solder flow and keeping the joints neat and clean.

[TAMHAN]

Hello,
sorry to ask stupid - that flux thing, so far, never grew to me. But I was at the random electronics shoppe today to buy something else, and they sold me the attached. Sadly, no thin solder - the thinnest they had was 0.6mm.

Is this recommended? Or would you suggest something else? I have a big Mouser order upcoming, so if someone wants to help me by giving an URL, I am even more thankful.

[Jwalling]

I would think that .6mm would work fine

Jay

[KE5FX]

If the chips are pin-compatible and it's simply a package mismatch, just bodge them in with (very short) jumper wires.  Chances are it'll work fine unless the design was extremely marginal to begin with.

[Jwalling]

If the chips are pin-compatible and it's simply a package mismatch, just bodge them in with (very short) jumper wires.  Chances are it'll work fine unless the design was extremely marginal to begin with.

Yeah, then post the results here:
https://www.eevblog.com/forum/repair/show-us-your-ugly-repair/ 

Jay

[timb]

So, are you upgrading to the 2M option by replacing the smaller SRAM chips with the bigger ones? I noticed on my 754C that the footprints have more pads than the installed memory chips use. So I assume it would be possible to replace them with the larger chips?

[Jwalling]

So, are you upgrading to the 2M option by replacing the smaller SRAM chips with the bigger ones? I noticed on my 754C that the footprints have more pads than the installed memory chips use. So I assume it would be possible to replace them with the larger chips?

He's just replacing one defective chip. I'm sure that upgrading it to 2M is theoretically possible on "C" scopes, as you say the pads are there, but not so practical given its age and value. 64 chips, yikes!

The "D" series of scopes above serial prefix B030XXX already have the correct SRAM parts for 2M. Over B040XXX, they went with synchronous burst-mode memory instead of SRAM. They can be upgraded to 2M easily as well.

Jay

Jay

[timb]

So, are you upgrading to the 2M option by replacing the smaller SRAM chips with the bigger ones? I noticed on my 754C that the footprints have more pads than the installed memory chips use. So I assume it would be possible to replace them with the larger chips?

He's just replacing one defective chip. I'm sure that upgrading it to 2M is theoretically possible on "C" scopes, as you say the pads are there, but not so practical given its age and value. 64 chips, yikes!

The "D" series of scopes above serial prefix B030XXX already have the correct SRAM parts for 2M. Over B040XXX, they went with synchronous burst-mode memory instead of SRAM. They can be upgraded to 2M easily as well.

Jay

Jay

Ah, that makes sense. I was just curious if it was possible; you couldn't pay me enough to do 64 chips by hand. Maybe with a hot air gun, but personally I wouldn't risk it.

[Jwalling]

Ah, that makes sense. I was just curious if it was possible; you couldn't pay me enough to do 64 chips by hand. Maybe with a hot air gun, but personally I wouldn't risk it.

You could pay me enough, I'm unemployed! If the price were right, I'd do it. I would hate every moment, what a tedious job that would be - 2048 solder joints to be made. 

Jay

[TAMHAN]

Didnt have time to reorder the new IC yet, so I cant report any news on the repair front. Plus, I opened up the 34401A I thought to have just ohms current source problems - nada, the entire DC Voltage area seems fooked, while the current source I spent hours repairing was fine and dandy all along.

But, in the mean time. AFAIK, there are two versions of the TEK754D planar in da haus: I cannot say anything about earlier versions (" ", A, B and C).

The one, which has BGA memory controllers and FOUR memory ICs is the 2M model and IMHO can not be upgraded. The one I have, picture attached, afaik is the 8M capable model. Why? Well. The same layout is also at the bottom of the acquisition board, which I had out recently but was too lazy to make a photo of its back. So we have 16 of these memory ICs per channel, which gives a whopping 8M sample space if I am not totally confused (which can be). If I recall correctly, upgrading involves the sending of some PITBULL GPIB command...

[Jwalling]

Didnt have time to reorder the new IC yet, so I cant report any news on the repair front. Plus, I opened up the 34401A I thought to have just ohms current source problems - nada, the entire DC Voltage area seems fooked, while the current source I spent hours repairing was fine and dandy all along.

But, in the mean time. AFAIK, there are two versions of the TEK754D planar in da haus: I cannot say anything about earlier versions (" ", A, B and C).

The one, which has BGA memory controllers and FOUR memory ICs is the 2M model and IMHO can not be upgraded. The one I have, picture attached, afaik is the 8M capable model. Why? Well. The same layout is also at the bottom of the acquisition board, which I had out recently but was too lazy to make a photo of its back. So we have 16 of these memory ICs per channel, which gives a whopping 8M sample space if I am not totally confused (which can be). If I recall correctly, upgrading involves the sending of some PITBULL GPIB command...

I think you're a bit confused. The 2M option was the maximum amount of memory ever offered in a TDS700C or TDS700D. This gives to 8MB of memory on one channel, or 2MB of memory on all four channels. See my post above. There is no option 8M

https://www.eevblog.com/forum/repair/tds754d-solder-advice-and-some-whining/msg1005193/#msg1005193

[TAMHAN]

Hello,
yes, that indeed was a misunderstanding. I meant with 2M, that a total of 2 Megapoints can be stored. And with 8M, that 8 Megapoints can be stored total.

Sorry for causing confusion here.

Tam

[Jwalling]

Hello,
yes, that indeed was a misunderstanding. I meant with 2M, that a total of 2 Megapoints can be stored. And with 8M, that 8 Megapoints can be stored total.

Sorry for causing confusion here.

Tam

You're still not getting it, or I'm misunderstanding you. There is no option 8M period. Option 2M is 8 megabytes of ACQ memory total. With option 2M you can store 8 million points on one channel only.

2M: Extends standard acquisition length from 50,000 points per channel to 2 million samples on three or four channels, 4 million samples on two channels, and 8 million samples on one channel.

Jay

[TheSteve]

Ah, that makes sense. I was just curious if it was possible; you couldn't pay me enough to do 64 chips by hand. Maybe with a hot air gun, but personally I wouldn't risk it.

I'd totally do it - crank the tunes and get to work. I would look at it as a challenge. I wouldn't want to do it day after day of course.
Years ago we took a bunch of 4MB simms and made them into 8MB for fun - same sort of deal, hundreds/thousands of PLCC solder connections(ram was way expensive then!).

[nctnico]

But on to the main problem. All I have is an ERSA i-con nano. I soldered some normal SMD stuff before, but never one of these SOIC ICs which have the pins bent under the main IC body.

Anyone, any ideas for me?

Put solder blobs on the pads, add flux, put IC on top and then use hot air to reflow.

[Jwalling]

Ah, that makes sense. I was just curious if it was possible; you couldn't pay me enough to do 64 chips by hand. Maybe with a hot air gun, but personally I wouldn't risk it.

I'd totally do it - crank the tunes and get to work.

For free?  It would be a challenge, and I know I could do it, but no way for free!
Would require at least three beers to steady my hands.
I would do it for $500 labor + parts. Any takers? 

Jay

[TheSteve]

Ah, that makes sense. I was just curious if it was possible; you couldn't pay me enough to do 64 chips by hand. Maybe with a hot air gun, but personally I wouldn't risk it.

I'd totally do it - crank the tunes and get to work.

For free?  It would be a challenge, and I know I could do it, but no way for free!
Would require at least three beers to steady my hands.
I would do it for $500 labor + parts. Any takers? 

Jay

I'd do it if it was my scope and I'd do it for a friend. It isn't something I'd really care to do for someone I don't know.

[Jwalling]

I'd do it if it was my scope and I'd do it for a friend.

Howdy Friend! 

[TAMHAN]

Hello,
sorry for asking dumb again: do you think the hot air process is preferrable to simply using the Ersa? I never worked with hotair so far...

Re the options: Tek says this:
50 K (Opt. 1 M: 500 K, Opt. 2 M: 8 M)

I miswrote it. I meant: if you have a planar, which looks like the one I pictured, you can enable 2M and get 8MPT for one channel. If your planar looks different, you can usually only enable Option 1M, giving 500K.

[Jwalling]

Hello,
sorry for asking dumb again: do you think the hot air process is preferrable to simply using the Ersa? I never worked with hotair so far...

Re the options: Tek says this:
50 K (Opt. 1 M: 500 K, Opt. 2 M: 8 M)

I miswrote it. I meant: if you have a planar, which looks like the one I pictured, you can enable 2M and get 8MPT for one channel. If your planar looks different, you can usually only enable Option 1M, giving 500K.

If the D model is B040XXX or higher, it has the burst mode memory and option 2M can be enabled.

Jay

[siggi]

Howdy Friend! 

[TheSteve]

I'd do it if it was my scope and I'd do it for a friend.

Howdy Friend! 

These are the same friends that helped me move when I bought a house and came over for a day to help dig a big hole and then mix/pour concrete for a new antenna tower base.