Tektronix THS7xxx Scope Hack/Teardown/Discussion - FW 1.16 found

[walter.aranda]

Dear jal1234, I am using version 3.0.0.0 (sending image), but I have reported freezing issues when recording data while connected to Windows 10 via a USB-RJ45 cable. However, this lag problem has been 'resolved' by using Windows 7 in a Virtualbox. Regarding the license issue you reported, I have read that Tektronix has released the license for domestic use (keycode: WSTRO618650124). I used that keycode, and it hasn't given me any problems related to licenses. Finally, regarding the baud rate, I also haven't encountered any issues, as long as the baud rate settings on both the PC and the oscilloscope match.

[jal1234]

Did you find a way to use anything other than 9600 baud in 3.0?  Some earlier versions had a faster option, but no option that I could find in 3.0.
Thanks,
Jon

[walter.aranda]

Although the recommended setting is 9600 I have not reported any problems when changing the baudrate, (I am using a modified FTDI USB-RJ45 cable, I can pass you my pin if you want), I did a test with the THS7XX HardCopy program which should run at 34800 in its configuration and I have had no problems.
Poniout color description:
| FTDI Cable | Description | New Color Pattern |
|------------|-------------|-------------------|
| 1 Brown    | CTS         | Black             |
| 2 Purple   | DSR         | Brown             |
| 3 Yellow   | RXD         | Red               |
| 4 Red      | GND         | Orange            |
| 5 Black    | GND         | Yellow            |
| 6 Orange   | TXD         | Green             |
| 7 Blue     | DTR         | Blue              |
| 8 Green    | RTS         | Purple            |

New order (tested):
   1        2       3           4           5         6        7          8
Purple | Blue | Brown | Orange | Green | Red | Purple | Black
  RTS    DTR     DSR      GND      TXD       RXD    RTS      CTS

[jal1234]

The Hard Copy program itself
 works at 34,800 baud, but the add-on analytic program doesn't work, so I'm left with the basic Hard Copy functions. So I went to the Tektronix program, which works at 9600. But while 2.x versions had speed options greater than 9600, the only option I see in 3.0 is 9600. The scope is capable, but the program appears not to be.

[mrprecision]

Hello,

I have tried the THS7XX HardCopy program and it worked with my scope. Also the post process with SignalAnalyzer worked fine for me. What are your issues?

Regards
MP

[jal1234]

It tries to connect, fails, and errors out. This is after I've verified the basic Hard Copy program works ok. Are you using 9600 or 34800 baud?
Jon

[mrprecision]

Which tool can not connect? Can you connect with THS7XX HardCopy and the the the scope display via remote? Can you set the scope to factory defaults and only change in the scope to 34800. The tool works only with 34800 and default settings.

[TERRA Operative]

Have you set the port on the PC to the same baud rate?

I'll have to experiment, but I seem to remember having to set that in Windows, maybe...

[jal1234]

It tries to connect, fails, and errors out. This is after I've verified the basic Hard Copy program works ok. Are you using 9600 or 34800 baud?
Jon

Hard Copy connects fine at 34800, and displays the screen. It can also do the quasi real time updates on screen. When I try to use the add on, then it won't connect.

Jon

[jal1234]

Update. Not sure what I did different, but after erroring out once, now Signal Analyzer connects. When it eventually opens, now I get this screen, sorry it's at 90°. When I click OK, I get just a blank screen.

Jon

[walter.aranda]

my friend, I just noticed that I have the same problem (virtualbox Win7 & FTDI cable)  as you when trying to communicate with baud rates higher than 9600. I tested with 19200 and 38400, and indeed, as soon as I change this setting on the scope, the hardware communication stops, and I even need to disconnect the cable to restore communication. I saw in the programming manual, Table 1-2, that it is recommended to leave the default configuration (I'll send an image); however, I don't understand why the options to increase the baud rate exist if this is not possible. Could it be due to some limitation from not using a pure serial cable? Could someone show if this problem occurs with an RS232-RJ45 cable for a PC's serial port? In any case, I can communicate at 9600, so I'm not too concerned about the transfer speed at the moment. Finally, I've noticed that I'm unable to do data logging from the scope to the PC—can you? Could you show me how to do it? If you wish, you can contact me privately. Best regards!

[mrprecision]

It seems that connection do SignalAnalyzer works. The free version is limited to 512 samples, so you need a licence key to use it.

[jal1234]

my friend, I just noticed that I have the same problem (virtualbox Win7 & FTDI cable)  as you when trying to communicate with baud rates higher than 9600. I tested with 19200 and 38400, and indeed, as soon as I change this setting on the scope, the hardware communication stops, and I even need to disconnect the cable to restore communication. I saw in the programming manual, Table 1-2, that it is recommended to leave the default configuration (I'll send an image); however, I don't understand why the options to increase the baud rate exist if this is not possible. Could it be due to some limitation from not using a pure serial cable? Could someone show if this problem occurs with an RS232-RJ45 cable for a PC's serial port? In any case, I can communicate at 9600, so I'm not too concerned about the transfer speed at the moment. Finally, I've noticed that I'm unable to do data logging from the scope to the PC—can you? Could you show me how to do it? If you wish, you can contact me privately. Best regards!

What version of Wavestar are you using? There was no option to change the baud rate in 3.0 that there was in earlier versions. Maybe the fact that Wavestar can't seem to communicate at that speed caused then to delete that option.
I haven't done much with Wavestar yet. I'll look and see what I can do when I get a chance.
Oh, and I'm using a serial cable with my Panasonic CF-19. No USB in the mix.

[jal1234]

It seems that connection do SignalAnalyzer works. The free version is limited to 512 samples, so you need a licence key to use it.

Oh, I see. The documentation to use SignalAnalyzer with the THS7xx is basically non-existant. I wish there was a trial period to try it before the limits kicked in. As it is, I'm hesitant to buy a license without making sure it actually works.
I've managed to get a couple of screens by fluke, I guess, a couple of times. Non-repeatable, and nothing useful was on screen.

[HAHSM]

Hello all, new member, my finding this discussion motivated me to join eevblog!

...I just recently picked up a THS720P as Parts/Not Working from ebay. Just the unit, no accessories or probes or anything. It seems to actually pretty much work so that's great, yes?!, have wanted one for a long time... 

It has the dc offset issue that others also have experienced, but not as great as some others have reported. Doing the cal often fixes that, we'll see if it fixes mine. I'm in the process of making a precision voltage reference and square wave generator so I can do the scope section cal myself.

The battery of course is dead - an original Tek NiCad from 1996...and even after attempting to charge for 24 hours it only showed 0.05vdc, dead. I made a 12v power cord and it powers up and self check & the simple cal functions all pass. Display is good, I get signals on both channels. DMM works fine. Seems I got lucky with this one!

I have parts on the way to make my own (2) NiMH battery packs from Amazon Basics 5000mah C's - I have a battery spot welder and nickel strips and PVC shrink wrap so should be easy to duplicate the OEM one...

So I've started digging into this scope a bit and have noticed a curious thing I haven't seen posted elsewhere...

But first, it's got 1.08 fw, would love to load 1.16 on it. I tend to think that 1.16 improves the cal function and is less prone to the dc offset issue. No way does my skill set include R&R'ing the necessary chip, but does include designing parts (in FreeCAD) and printing 3D parts. Seems like I could make the plug adapter that plugs into the main board similar to the factory setup, but don't know how to operate the programmer to actually upload the firmware. Last time I programmed IC's it was EEPROMS and EPROMS, remember those?  I have read all the related posts in this thread so know the extents folks have gone to - to try to upload the firmware. Perhaps I could team up with others to make a standard cable setup for firmware updates, if folks are still interested, I sure am.

OK - curious thing - the Service Manual says to remove the cal jumper to enable the calibration function, I did this (not the cal yet though) and sure enough saw the cal functions in the menus afterwards. Curious thing though is that it's not a 2 pin header that the jumper goes on but a 3 pin header with 2 possible jumper positions - two leftmost pins or the two rightmost pins. Originally the jumper was on the two leftmost pins for normal functionality. The Service Manual doesn't point out to make sure to put the jumper back on the two leftmost pins, nor does it say what happens if the jumper is placed on the two rightmost pins. So of course I had to try it. Bizarre!  On powerup it immediately goes into some loop that appears to be strange screen diagnostics and I couldn't find a way to exit the loop - all keys and even the power button was ignored. So I pulled the 12v supply, thinking this would cause immediate power off since the battery is flat dead -NOT- the scope kept running the strange screen stuff with no power source - off of what power source? I wanted to put a quick end to this since it wasn't clear what was going on, and I want this scope to survive, so I popped the battery pack out - screen stuff still running - and immediate power off. Turns out putting the jumper in the rightmost position made the scope revive and start charging the battery again (in addition to whatever was going on with the screen, and who knows what else in the background.  So the scope runs now off the 30 year old previously flat dead battery...and with the jumper back over the leftmost pins the scope operates normally again.

Anyone else experience this? I haven't seen any other posts about the 3 pin header for the cal jumper, anyone else know about this and what the rightmost position is intended for?   I didn't see anything in the Service Manual (haven't checked the schematics yet).

Also I'm curious how others have made their own cal circuits to do a diy calibration (for those of us who don't have a nice bench full of signal generators and precision power supplies and scope calibrators and such.
I do have several trusted Fluke DMM's and an older reasonably accurate 60mhz analog scope so I should be able to hack together some reasonably capable diy cal equipment.

For the precision voltage references I'll be using a REF10EZ 10 volt precision reference, then simple resistor voltage dividers for the 3 needed +-voltages, adjusted via precision trimmer resistors.

For the square wave generator I thought I'd try a simple rc oscillator into a 74HC14 hex inverter with Tr and Tf avg of 7ns - so well under the 20ns rise time spec'd in the Service Manual. I'll do my best to verify accuracy of these before using them to do the cal.

I look forward to replies 😎

[TERRA Operative]

Check out the tekwiki page for info on these units, including the pod needed for firmware upgrades. If someone were to emulate that with a raspberry pi or other suitable device, it would make for much easier firmware upgrades.
The motherboard looks simple to replicate from the schematics, but the Motorola M68332BCCDI 'DIBUG' debug module is unobtanium...
I haven't been able to find an exact match for the plug though, it appears to be a SCSI or similar type of thing.

https://w140.com/tekwiki/wiki/THS720

https://w140.com/tekwiki/wiki/067-1436-00


According to the schematics on Page 10 (Grid Ref A1) the 3-pin cal header is also some sort of test function, probably used during manufacture. I wonder if it spits something out over the serial cable?

For cal, I use some Advantest and Yokogawa AC and DC references, a bunch of Vishay precision resistors, and a HP 33120A Function Generator (Specified as <20ns, so juussstt good enough).


If you make the squarewave gen, share the design (Maybe in a new thread), I'd be interested in making one too.

[HAHSM]

Check out the tekwiki page for info on these units, including the pod needed for firmware upgrades. If someone were to emulate that with a raspberry pi or other suitable device, it would make for much easier firmware upgrades.
The motherboard looks simple to replicate from the schematics, but the Motorola M68332BCCDI 'DIBUG' debug module is unobtanium...
I haven't been able to find an exact match for the plug though, it appears to be a SCSI or similar type of thing.

https://w140.com/tekwiki/wiki/THS720

https://w140.com/tekwiki/wiki/067-1436-00


According to the schematics on Page 10 (Grid Ref A1) the 3-pin cal header is also some sort of test function, probably used during manufacture. I wonder if it spits something out over the serial cable?

For cal, I use some Advantest and Yokogawa AC and DC references, a bunch of Vishay precision resistors, and a HP 33120A Function Generator (Specified as <20ns, so juussstt good enough).


If you make the squarewave gen, share the design (Maybe in a new thread), I'd be interested in making one too.

Hey thanks for the reply - in the Pod manual is a replacable parts list with the Tek part number for the adapter cable (with plug) to connect to the THS - part 671-3682-XX, doing a search for that 😎 ya good luck, right! haha.

I see that jumper in the schematics now - thanks for pointing to the location in the schematics, it seems to do some interesting things like (as you said) indicate some sort of diag mode, maybe the diag mode is the one in conjunction with the Pod?...

If I get my hands on a THS mainboard, or just a cheap parts only THS, I'll see if there is an AMP number on the that plug. Seems like there would be...

I'm waiting on some of the parts from Mouser for my diy cal circuits, if they work out & the square wave gen is nice and stable/no or low jitter then I'll (as you suggested) post a new thread with what I came up with - I'll also send you a msg with the new thread.

I've built my two battery packs, they work great, charge overnight, just get very mildly warm so not overcharged.

I made a RJ45 to DB9F serial cable and tested it as working with hardware flowcontrol.

So this THS720P that I bought as "For Parts/Not Working" is fully functional, only issues now are doing a firmware update and the DC offset in both channels issue (that I'm hoping the diy cal will fix).

More soon...

[HAHSM]

According to the schematics on Page 10 (Grid Ref A1) the 3-pin cal header is also some sort of test function, probably used during manufacture. I wonder if it spits something out over the serial cable?

When I do my cal procedure (& have the scope opened up again) I'll try the jumper in the right/Diag location again & monitor the serial port (now that I have a serial cable) and see if it does in fact start spitting stuff out...

[kblue]

Tried to get rid of the little but noticeable offset with a cal, but the step +7.000 VDC always fails, regardless what source i use and according to my bench-multimeters, which are connected during the cal, its 7.00012 +/- VDC and yes, the 50 Ohm feedthrough terminator is removed already. Bummer.

Well guess i need to live with the offset. The scope is running older firmware 1.11.

/edit: Nevermind, it was my fault! I had my Banana-to-BNC connector reversed so +7 was -7 for the scope. <self-slap>
/edit2: Kinda funny ... i did the cal with the original power supply connected. Back to batteries, there's some "tiniest of offset" back again and gone with power-supply connected (w/o batteries).

[TERRA Operative]

There are four optoisolators that tend to drift. They can be directly replaced with Broadcom/Avago part number HCNR201 (any of the -300E, -500E, -350E, or -550E variants will work) which are still readily available.
I just bought 4 this week to keep on hand for my THS720P, just in case I need them.

[kblue]

Hello again!

So far the unit works perfectly fine. Just the old NiCd batteries are quite dead. Some people straight up replaced the old ones with NiMh in a tube, others say NiMh without modifications is a bad idea.

My CCFL works fine too and is very bright, still I wonder if some sort of LED replacement would give me more runtime.

[TERRA Operative]

Ni-MH is fine to use. I have rebuilt all my packs using 4000mAh GP brand C-cells and they work great.

An LED conversion will increase runtime too, you save quite a bit of power by ditching the CCFL for LED's