Upgraded TDS340 to TDS360 200Mhz @ 1GSa/s. For Any One Interested .Plus TDS380

[Labrat101]

Hi,
First I know this subject came up about 3 years ago and there was never any conclusion how it turned
out.
Well being bored and had too much time on Hand etc.
I decided what the heck I will revamp and old TDS340 and give it a new lease of Life and Kill a few hours.
 The whole project took about a day including research and studying the service manual .
And doing it with minimal parts OK what I had at hand .
The only Parts I did not change in the List were the four 270nH inductors as I didn't have and guessed
if all went to plan I would order them .
 L203 ,L204,,L205 & L206 Need to be replaced for 27nH  100Mhz Q ) 
Changed 7 resistors  , 
All  1%
R208 2.21k changed to 2k  ( important )
R240 10k  (added)
R241 3.3k (added)
R246  222   to 10
R247  222   to 10
R248  222   to 10
R249  222   to 10
Removed 1 cap C125 (AE1 15pf) located under the shield near C219 
Removed 2 bridge links R651 & R653   0.0   (green)

 The exact location of the parts are Marked on the Photo,s  as coloured dots.

Next Plug it in   
Long Pause .
It Booted up just fine with the new screen TDS360   
It then needs to be recalibrated  .. Utility .. Cal .. wait for about 5 minutes 
O yer all passed with flying colours . 

If any one else is interested in this nice easy upgrade Please Enjoy .
The scope must be functional as A TDS340 and not showing any error . Before upgrading .
 
This project is, as is . I will not be responsible if you get it wrong .

 If this Helps someone please just say THANKS 

If you Do need Help Please feel free to Ask .
 Also I know How to do a full Calibration of these TDS easy way If you should loose
 the cal . and get Stuck .  You will Require a Calibrated Meter Accurate to 4 Digits ie uv  (Updated )
 Read the entire Blog 

[Labrat101]

Few more Pictures

Update : If someone has the newer firmware 1.05 that would make a great addition to this.
I have been looking for it without any joy so far .
So if some one could throw some light on where to find it .
   would be more then welcomed

[Labrat101]

Just another update .
That might be required for fine tuning if there is a small problem with the wave form .
 varies from device to device .
 
  C230 & C231 are normally fitted on the TDS360 By Default :: 
CERAMICMLC;1PF,+/–0.25PF,50V
     

[james_s]

Also I know How to do a full Calibration of these TDS easy way If you should loose
 the cal . and get Stuck .

I just noticed this part. Between a friend and myself we have three of these things that have all lost their calibration. I've been successful in calibrating the voltage and trigger but I have not managed to get the timing calibrated, and I'm not spending $20k on the instrument they suggest using to calibrate it.

[cheapskate]

The TDS380 is also really similar, I wonder if it would be possible to upgrade further and unlock 2GS/s mode. It seems that the four FISO inductors of the 340/360 are replaced with some high speed NPN transistors so probably not.

The firmware is likely stored in U602, U603, and U607 with calibration values stores by U605. The address and data busses for those chips are present at the GPIB/Parallel/Serial options board but aren't connected in their entirety to the GPIB or UART controllers. Unless someone figures out how to boot the MC68331 CPU from an external source or the firmware includes some secret features, I don't think it is possible to extract the firmware without desoldering the flash memory. Perhaps OC3, OC4, and OC5 on J608 are some sort of debug UART?

[james_s]

I know with certainty that the calibration values are stored in the Dallas NVRAM because dead batteries is the reason the scopes I have need calibration.

[Labrat101]

I know with certainty that the calibration values are stored in the Dallas NVRAM because dead batteries is the reason the scopes I have need calibration.

Hi James ,
First you don't need to spend $20K  unless you feel like giving me a BD present.  .
If you Have lost the CaL . The battery must be replace if not already done so. as they some times
die short circuit 

If you managed to do all the voltage calibrations OK . Though the 50ohm terminator . you guessed that
when it ask for 1v you have to set the function gen to 2v so the scope see 1v @ 50 (Example)
When it ask for the fast rise pulse this also must be doubled so the scope see 1v pulse @ 50 . (example)
If using a UTG 962 set to 50ohm load the out put voltage is doubled .corrected for 50
  put a bench meter on the function gen on the DC 1.0000v and measured what the voltage error is
from the function Gen (example) set dc level to 1.0000v  Meter read 1.0010 . so minus the error on the
function gen . So when set to 2.0000 volt  - 0.0020 to correct . (use a RG cable with BNC to Banana )
Don't measure with single jumper wires .
then set function gen to pulse and the corrected voltage and all should be good.
The TDS has built in ref voltages & frequencies these have to be matched before it will load them back
into NVRam again . 
If this still fails then you may have a few out of tolerance caps . See photo above the Blue caps were
out on my one. these were  caps I replaced 1% mylar Good enough and this solved that .
I guess you don't need NASA spec's . After Cal I checked with a Leo Bodnar GPSDO .. less than 1% error.

Or you can raise or lower the signal voltage by just a few uv will do it as well to compensate for wear and tear.

The one Most important thing that I found that this scope is very fussy on noisy connections during Cal setup
the connections MUST Be 100% good connections including the grounding.
As any voltage drop on the cable will result as a fail    Did it to me as well .

The Battery MUST be replaced or you will be recaling every other day .
or never unplug it from the mains .
If you use the UTG962 Set to 50ohm Load the voltage will be correct Set@987mv =1.9990v
 Scope sees after terminator 1.0005v Adjust to match cable losses   (updated typo)
Did this Help or did I miss understand something with the Question

[Labrat101]

The TDS380 is also really similar, I wonder if it would be possible to upgrade further and unlock 2GS/s mode. It seems that the four FISO inductors of the 340/360 are replaced with some high speed NPN transistors so probably not.

The firmware is likely stored in U602, U603, and U607 with calibration values stores by U605. The address and data busses for those chips are present at the GPIB/Parallel/Serial options board but aren't connected in their entirety to the GPIB or UART controllers. Unless someone figures out how to boot the MC68331 CPU from an external source or the firmware includes some secret features, I don't think it is possible to extract the firmware without desoldering the flash memory. Perhaps OC3, OC4, and OC5 on J608 are some sort of debug UART?

Update  The NVRAM (U605) consists of a single nonvolatile memory IC. This RAM
provides storage of front-panel settings, waveforms, and
calibration constants
TDS 360 to 380  does not need a firmware update  Only the Sampling chip is changed and 4 buffer transistors the inductors are removed completely . and a few links & caps 
I am not going into this .  There is a flash point and its Not via GPIB .
 Only If the newer chip was already installed on the 360.  and its the B version MB then you have a
Project . other wise there will have to be a few additional jump wires  .
I am already pushing my upgraded 340/360 to 1.2Ghz BW which is maxing the sampling chip.
with some custom made caps .

[Labrat101]

Don't forget to remove the Cal Jumper J609  first ..

I have made a bad drawing of the connection .  (Not as good as Dave's Cad)
All lead Must be the same Length 50cm MAX
The Terminator at the scope Must be a Straight Though 50 Terminator .
All cables Must be high quality 50  and no loose connections
A Banana Plug to BNC required for the Meter Fluke or similar

 
This is the only correct way if you don't want the failed Message on xtrigger

[Labrat101]

Cal update The Trigger Cal when the scope ask for it
is done as Per Photo.   
Please note checking all connection this way and you will know that all is good 
Also you can check final voltages that the scope SEEs
Will help those who are having trouble getting the scope to recognise the voltages etc Asked for 
IE Connecting the Fluke instead of the scope to the 50ohm load . and adjust your voltage ref to the correct voltage
on the Fluke  when your happy with your setup and all is correct . The calibration will go easier .

Good Luck .  And If this Helps you Please enjoy . and Say thanks .   
 If it made you Laugh that is also Good   

Sorry I must be the worst teacher on the planet . I just Know how to do . But explaining how I did it is
some thing else .        Still confused . 
  Built My first all Tube scope when I was 10 yrs Old  (Early 60s )

[james_s]

Cool, thanks for the info. Yes I replaced the battery no problem, that's when I ran into the issue trying to calibrate the timing. So you did it using a cheap Chinese function generator? I was becoming convinced that something much more exotic was required. I have a 33120A which is supposed to do better than 20ns rise/fall, I don't know offhand how that compares. Seems like the Tek manual specified something under 10ns for calibration but maybe it doesn't actually matter. I'll have to take another crack at this later.

[Labrat101]

Cool, thanks for the info. Yes I replaced the battery no problem, that's when I ran into the issue trying to calibrate the timing. So you did it using a cheap Chinese function generator? I was becoming convinced that something much more exotic was required. I have a 33120A which is supposed to do better than 20ns rise/fall, I don't know offhand how that compares. Seems like the Tek manual specified something under 10ns for calibration but maybe it doesn't actually matter. I'll have to take another crack at this later.

I sent you a PM before I read this .
 I was sure I used 1khz @ less than 10ns rise .  as I said in the PM I think the manual has a TYPO .
 Only Thing you need is an accurate voltmeter that can read uv  (4 place of decimal )
 Remember 25 years ago there was no such thing as cheap Chinese function generators 

  My new Project is HP54645A  .. Hehe ..I have got my sticky little hands on the original A1 Blueprint ..

[cheapskate]

The TDS300 programming manual includes an undocumented "PITBULL" command. On the TDS500+ "PITBULL" is used as a password to allow unrestricted memory reads and writes. I wonder if this command is somehow similar. It is only mentioned in the reserved words listing, there is no associated documentation. I don't have a null modem adapter for the serial port yet so I haven't tried it out.

[Labrat101]

The TDS300 programming manual includes an undocumented "PITBULL" command. On the TDS500+ "PITBULL" is used as a password to allow unrestricted memory reads and writes. I wonder if this command is somehow similar. It is only mentioned in the reserved words listing, there is no associated documentation. I don't have a null modem adapter for the serial port yet so I haven't tried it out.

Hi  Cheapskate I found this in My Archive .
I did not Pull this and I can NOT verify Firmware version etc .
If you want to open it and mess with the coding Enjoy .
 I Do NOT work with Software . I am a Hardware engineer . I let others play with the BIN codes
 So Use at your OWN risk . If you do figure out an easy way to update the software Lets us Know
how you got on  .
But I am a big believer in If it aint broke Don't Fix it

PS . I am sure of One thing you don't have to mess with the firmware to take it
up to 2Ghz Sampling . as the the input section on these scopes are 400Mhz
rated and the current firmware recognises the faster speeds if the sampling chip
is replaced . If your scope was original TDS360 the transistor place footprints
are under the 4 inductors . If it was Made AS a TDS360 they should be already
installed . And just follow the custom sheet I put Above earlier
 #7 on: January 13, 2021,
 Plus Look on the schematic Diagram Last section it will show the Links to be removed or added . 
Have Fun Enjoy.

[Labrat101]

Just an Update
This is the Cal Frequency Screen . as per the Manual
The Cal jumper Removed ..
Recall Factory setup first  .. Must
Please note that if the settings are not exact as shown the screen on the scope Will not
Snap to this  . As all the references Must match the Internal setting
The Square wave is  NEGATIVE    (1ms)  1Khz @ -200mv
 Trigger is set to -102mv .. MUST
 Time base 200 MS/s     250 ns   .. MUST
 Volt / Divide 50mv  ..  MUST
All these setting have to be as is .

[kozard]

I have been sending messages back and forth to the thread opener as I made the modifications to a "For Parts" TDS 340 I bought from eBay. The thread opener suggested that I post.

Originally I first bought a TDS 380 from eBay and it arrived with a blown attenuator. As far as I can tell that is not practically fixable unless you sacrifice another TDS 380 and are really capable of desoldering the ceramic hybrid. [The preamp IC is damaged and drags down the DC offset input to -7V and saturates the preamp output (clips).]

I compared photos from the non-functional TDS 380 (400 MHz) to the "for parts" TDS 340 (100 MHz) and TDS 360 (200 MHz). Originally I bought the TDS 340 to salvage the attenuator as a spare part to repair the TDS 380. However I found that the TDS 340 and TDS 360 have a different attenuator inside. The part number on the preamp IC is different. As a result my plan changed to modifying the TDS 340 to perform as close to the TDS 360 (or even better).

Based upon the attached TDS 380 photos (with 10 Ohm resistors and no bandwidth limiting filter/inductors) I decided to use the same values on the TDS 340 upgrade to achieve what I hope will be the maximum possible bandwidth upgrade.

Note that the TDS 380 also has four SOT23 marked R2 which I believe are SMD versions of the 2N3906. I don't plan to make a tiny PCB board with four 3906 buffers but I guess that is possible if you want to get the absolute maximum bandwidth.



I followed the instructions in the first post however I made a couple of modifications:
I did not install 27nH inductors in place of the 270nH inductors. Instead I used 0 ohm 1206 resistors. (Since the TDS 380 has no bandwidth limiting filter/inductors.)
I did not have a 2k 1206 resistor so I soldered a 22k on top of the existing 2.2k to achieve 2k equivalent.



The TDS 340 now boots up with a display that says "TDS 360". I ran signal path compensation and it passes.



Photos are in the following posts.

[whatboy]

Maybe  your photos are too big, try resizing to lower resolution..

[Labrat101]

I used PNG for most of my clips here .   
 I believe its Snip in windows  or Screenshot in Linux with Grab area .
I only use Linux .. Anti Bill thing

[kozard]

These are pictures of the TDS 340 after the upgrade.

[kozard]

Here are the TDS 380 photos.

In the attenuator photo you can see that the preamp chip is not the same chip that is in the TDS 340 and 360 attenuators. Originally I bought the "For Parts" TDS 340 to fix the TDS 380 that has a blown attenuator. [The preamp IC is damaged and drags down the DC offset input to -7V and saturates the preamp output (clips).] If anyone has any ideas on how to fix it please let me know. I think it is perhaps too much even for MacGyver since any little replacement PCB for the preamp chip needs to do the DC offsets (H Position), HF adjust, three attenuation levels, balanced output and be close enough to the original hybrid to pass calibration and signal path compensation. As far as I know the preamp chip and the hybrid are both unavailable. I have also read that removing the brittle hybrid is not easy.

In the zoomed in photo you can see the R2 (MMBT3906?) buffers which presumably are needed to get the highest bandwidth (400 MHz) in the TDS 380.

In the TDS 380 the resistors are 10 Ohms and there are no bandwidth limiting filters (the inductors which are 270nH in TDS 340 and 27nH in the TDS 360). Presumably it is the inductor combined with the input capacitance of the IC is what forms the bandwidth limiting filter. That is why I used zero ohm 1206 resistors in place of the inductors and 10 Ohm resistors when I upgraded my TDS 340.

[Labrat101]

Nice . I am glade every thing went well . Saved another TDS . The photos look good .
I notice your Dallas chip has still got the original battery . you may have to replace it in the near future .

[kozard]

If anyone would like to see what a real TDS 360 looks like inside I have attached photos from a scrap TDS 360 that I bought for spare parts.

As you can see the TDS 380 PCB is not the same as the TDS 340 and TDS 360 PCB.

The TDS 340 and TDS 360 share the same PCB and appear to be only slight BOM/Bill-Of-Materials (parts list) changes.

The TDS 340 and TDS 360 use the same attenuator. The TDS 380 uses a different attenuator with a different part number on the preamp chip that is soldered to the ceramic attenuator hybrid.



In this post you can see the 27nH inductors that are used in the bandwidth limiting filter (200 MHz) of the TDS 360. I used zero ohm jumpers in the hope of getting a bandwidth between that of the TDS 360 (200 MHz) and the TDS 380 (400 MHz).

[kozard]

Yes, I do need to replace the DS1644. I just bought a DS1744 off of eBay from a US seller and it has a 2020 date code. I am hoping that the DS1744 will work in the TDS 340/360 as a replacement for the DS1644.

I am planning to use ChipQuik Tin-Bismuth low temperature solder to remove the existing DS1644. I don't know how to backup the DS1644 inside the scope. I am hoping to remove the DS1644 without too much heat and I hope I can still read it afterward. I hope it will not be corrupted or destroyed.

[See the attached picture of the Arduino Mega 2560 programmer with the DS1744 installed.]

If someone else needs a DS1644/DS1744 reader/writer mine is based upon the following Arduino EPROM programmer. [It is very easy to make but a little slow. Don't forget to put a pull up on the Write Enable pin! In my case when I try to read the DS1644 removed from the scope I will have the Write Enable hardwired to the Vcc pin just to prevent that chance of corruption.]

Here is the link for the Arduino EPROM programmer that I found online: http://danceswithferrets.org/geekblog/?page_id=903



Finally for completeness I have attached pictures of the unmodified TDS 340 just in case someone needs to see an unmodified TDS 340. The inductors are 270nH I believe.

[Labrat101]

If you copy the calibration from the Dallas after upgrade 340 to 360 I have a feeling that the cal will be out and you will still
have to do a full re calibration . I had to do a calibration after I replaced the battery with the drimmel out job .
Then again after the upgrade I found my one was out again .  Re ran a full calibration again . Maybe Luck of the draw . would be
interesting if the Cal holds true after upgrade as the new ranges won't have a cal point .

UpDate Edit :: What I mean is that the Calibration flag is still set as OK . when 340 mode now after the Upgrade to 360 the Cal flag wont necessary be changed and now with Bandwidth changed etc . I would be inclined to check / verify other wise you maybe doing some tests on a future project
 and be led down a Rabbit hole look for false error . I would personally feel safer re: doing a full calibration or at least a partial .
Yes I know its a pain in the  Ass but could save a future      Cal!Out!!   
   

[kozard]

Finally made progress on fixing the real TDS 380. [Funny how that happens immediately after I upgrade a "For Parts" TDS 340 to a TDS 360.]

There is (hopefully) some useful information here for others.

For example, in the photos you will see that I replaced the DS1644 with a DS1744. I can confirm that this works. (I could not find an answer to that question when I started.) I expect it will also work in the TDS 340 and TDS 360 in place of the DS1644.

Another important fact is that assuming your scope is not damaged (and thus can be recalibrated) you can install a blank DS1744 (or blank DS1644) into the TDS 380 and start the calibration. You don't need to save the data from the old NVRAM. I would prefer to save the data myself but the TDS 380 arrived in terrible condition. (See the last photo for what I received from eBay!) Note that attenuator damage (such as burned resistors on the ceramic hybrid or over voltage damage to the MAXTEK preamp chip on the ceramic hybrid) will cause calibration attempts to fail.



Next I tried to build a number of voltage calibrators and the Calibration Voltage Reference always failed. (Error message "Cal error, CVR source gain".)

Here are some hints to make it work: Listen CAREFULLY and observe the timing between the individual voltage calibration steps. If it quickly moves to the next step then you know that it has already failed. Make a note of that step and improve your calibration voltage source for that voltage step! Otherwise you will go through all eight steps and not know where the failure occurred. In my case it passed with a battery powered home-made calibrator with a battery powered meter attached. (Not necessarily a good idea to leave the meter leads (antenna) attached but it worked in this case.)



Now you can build a relatively low-cost calibrator. I finally got it to pass when I used a battery (3x18650) and op-amp buffers to drive the 50 Ohm terminator. Each op-amp buffer has an RC filter at the input using a 0.47uF polypropylene capacitor. Note do not use electrolytics for the noise filters due to leakage current. The final step is 15.0mV which isn't very much.

I did not have the ideal rail to rail op-amps available. I really needed low offset voltage and low input bias current precision op-amps that are rail to rail at least to the negative rail on input and output. I did not have those so I had to be extra careful about using low resistance dividers due to the problems cause by the input bias currents. I don't recommend that you use the op-amps I used. If you must build your own calibrator try to make your life easier with a good precision reference and good quality precision rail to rail op-amps that are low input bias current. I recommend battery power.



I used 3x18650 followed by LM7805 followed by a TL431 to generate 2.5V Vref. (The TL431 is not great but that is what I had. Try to get something better.)
Then I used a resistive divider, RC filter and op-amp buffer to generate 1.5V from the 2.5V Vref.
Then I did the same with a resistive divider, RC filter and op-amp buffer to generate 0.3V from the 2.5V Vref.

I could not successfully generate a sufficiently stable 80mV and 15mV from the 2.5V Vref.

So I used the buffered 0.3V with a low impedance resistive divider, RC filter and op-amp buffer to generate 80.0mV.

Finally I used the buffered 80.0mV with a low impedance resistive divider, RC filter and op-amp buffer to generate 15.0mV.



I could not get it to work with relatively low noise HP supplies, resistive dividers and RC filters. I suspect it was a combination of going to battery power and op-amp buffers (with RC noise filters) which made it finally work.

I am not saying my calibrator is the way to do it. Just that it worked and I was able to build it without buying anything. I am sure there are much better ways to do this, especially if you are willing to buy more expensive parts and/or equipment.



I have attached pictures. Note the last two pictures. The second to last is this TDS 380 as it is today. The last picture is the TDS 380 as it arrived from eBay! (With apocalyptic damage done inside by someone else.)



I placed this here in this thread because I believe it might possibly help others calibrate a TDS 340 upgraded to TDS 360, for example. Such as the "For Parts" TDS 340 that I just upgraded to TDS 360.