Author Topic: Keithley 2400 SourceMeter review and teardown  (Read 36746 times)

0 Members and 2 Guests are viewing this topic.

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #75 on: February 29, 2020, 04:49:15 pm »
Quote
U660 AD7849 is hot and not changing the current setting.

U660 is voltage (not current) control.

Although fried DAC would explain observables nicely.
 
The following users thanked this post: kawal

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #76 on: February 29, 2020, 04:55:50 pm »
Yes U660 is removed and the situation is not changed apart from NVdac showing +12V everything else looks the same :palm:
TP213 was -0.75V
Getting expensive

Either way why is there -240V on the output when teh output is not enbaled.
« Last Edit: February 29, 2020, 04:58:29 pm by kawal »
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #77 on: February 29, 2020, 05:15:19 pm »
Quote
U660 is removed

Don't rush into things.  :)

TP213 (THE control) can be driven down either by voltage drive: TP205 near U211 (not TP206!) OR by current drive: TP207 (near U213)

Please check those voltages. (to figure out is it current control that screws things up or is it voltage control)
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #78 on: February 29, 2020, 05:30:14 pm »
TP205 = 0.566V
TP207 = -7.4V
Openloop
Do you have a schematic of this unit ?
« Last Edit: February 29, 2020, 05:56:01 pm by kawal »
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #79 on: February 29, 2020, 06:03:55 pm »
Quote
Do you have a schematic of this unit ?

Nope, just bad experience.  ;D

Quote
TP207 = -7.4V

So the current control is bad...

Let's check the summing point where feedback is added to (current) DAC's output:

R238 pin 1: DAC
R238 pin 2: DAC + feedback
R238 pin 3: feedback

What are the voltages on those pins?
« Last Edit: February 29, 2020, 06:16:52 pm by openloop »
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #80 on: February 29, 2020, 06:30:35 pm »
openloop

DAC U660 is back in - still runs hot - super hot
When set to 10V out the NVDAC pin shows -5.00V
DAC U661 is cold and  output  on pin NIDAC now remains 40mv  no matter what current is set - so I killed something
I set ouptut to 10mA
With these conditions I get the following readings
OP AMP out -1.2V
TP213 -2.3V
TP212 0.36V
TP211 -1.67V
TP210 +7.8V
TP209 +12.7V
TP205 +12V
TP207 -0.44V
TP230 -13.9V
TP231 -0.83V

 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #81 on: February 29, 2020, 06:57:28 pm »
Quote
Do you have a schematic of this unit ?

Nope, just bad experience.  ;D

Quote
TP207 = -7.4V

So the current control is bad...

Let's check the summing point where feedback is added to (current) DAC's output:

R238 pin 1: DAC
R238 pin 2: DAC + feedback
R238 pin 3: feedback

What are the voltages on those pins?

R238
pin 1 = -43mv
pin 2 = -0.833V
pin 3 = -0.833V
pin 4 = -0.833V
pin 5 =  -43mV

 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #82 on: February 29, 2020, 07:18:29 pm »
Quote
TP213 -2.3V
TP211 -1.67V
TP207 -0.44V

That does not compute. Can you recheck those three? Value at TP207 seems wrong.

And while at it, check voltages at TP205, TP206, TP208
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #83 on: March 01, 2020, 12:05:16 am »
Quote
TP213 -2.3V
TP211 -1.67V
TP207 -0.44V

That does not compute. Can you recheck those three? Value at TP207 seems wrong.

And while at it, check voltages at TP205, TP206, TP208

OK TP207 was wrong its +0.44V

Rechecked the Resistor network as well
R238
pin 1 =  -43.3mV
pin 2 =  -0.438V
pin 3 =  -0.833V
pin 4 =  -0.833V
pin 5 =  -0.395V
pin 6 =  +43.3mV

OP AMP out -1.2V
TP214 -2.3V
TP213 -0.4V (corrected)
TP212 +0.4V
TP211 -1.6V
TP210 +10.3V
TP209 +10.2V
TP205 +9.4V
TP207 +0.44V
TP230 -13.9V
TP231 -0.833V

While messing with the unit I set it to current mode and complacence voltage to 2.100V and the meter measured 2.10018V I checked the output and it was right no the money (2.10018V measure with Keithley 2700). The current was set to 10ma but the DAC must be dead as it always Measures  43mV on NIDAC . Current reading on the meter is 0.833mA and the  output stage is not outputting any current at least based on the measurement across   R562 R563 being 0.000mV
Tried it with Compliance Voltage of 21.0v and that works as well.
But current is always reading 0.833mA same as pins 2, 3 off R238. So the current feedback must be trying to correct for its reading.

U227 feeds directly to the pin IFB which I think is Current Feedback.
And this is the issues there is a large voltage  on the indeed of this chip. The chip is a AMP03G  which makes sense.
Now to figure out where this voltage difference is coming from there is a ton of switching going on in this area.
 
In voltage mode i get -240V all the time even when output is off.  So something is working.


Updated firmware from version C11 to C32  while i was inside. Used images from Xdevs and  TL866II  plus programmer.
The memory was TMS28F020 and I saved C11 just in case.
So the measurement side is working and the output side is working .
NV dac still crazy hot probably needs to be changed - but outputting correct values.
 

« Last Edit: March 01, 2020, 02:35:12 pm by kawal »
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #84 on: March 01, 2020, 01:28:57 pm »
You provided different values for TP213. It was both -2.3V and +0.4V
Did you mix-up TP214 and 213?
TP213 cannot be -2.3V - it's clamped.
Man, focus!

+0.4V at TP213 is good so control input is correct - it tries to pull the output up.

If TP214 is indeed -2.3V then the suspicion lands on U236 (and its neighborhood) which is the last buffer between TP213 and TP214.
Check voltages on its pins.
pin #3 must be the same as TP213, pin #2 should be at floating ground (FCOM), #4 and #7: 15V power, pin #6 output - in this situation it should be driven to its positive rail.







« Last Edit: March 01, 2020, 01:41:34 pm by openloop »
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #85 on: March 01, 2020, 01:43:10 pm »
You provided different values for TP213. It was both -2.3V and +0.4V
Did you mix-up TP214 and 213?
TP213 cannot be -2.3V - it's clamped.
Man, focus!

+0.4V at TP213 is good so control input is correct - it tries to pull the output up.

If TP214 is indeed -2.3V then the suspicion lands on U236 (and its neighborhood) which is the last buffer between TP213 and TP214.
Check voltages on its pins.
pin #3 must be the same as TP213, pin #2 should be at floating ground, #4 and #7: 15V power, pin #6 output - in this situation it should be driven to its positive rail.

Yes got TP213 mixed up at the very start adn continued to name it incorrectly.  TP214 is EAMP and that is -2.3V in voltage mode but it is absolutely fine in current mode (around 0V).
Pretty sure the issue is not U236 as it works as should in current mode
Either way DACs are hot and current  dac is dead along with the current sense loop showing something that is not there.
I think at the end i need to figure out the current sense loop around the resistors at the input and the switching of these resistors. There is a voltage imbalance there.
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #86 on: March 01, 2020, 02:03:29 pm »
 |O

So what's on TP213 in voltage mode?
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #87 on: March 01, 2020, 02:06:22 pm »
TP213 -0.4V.....-0.76V  (must have switched leads at some point as it seems to always negative)


OK  in Voltage mode :
   TP213 in voltage mode is -0.76V  in lower current ranges and -0.4V in higher current ranges
   TP214 is -2.3V
In Current mode
TP213 440uV
TP214 70uV


« Last Edit: March 01, 2020, 02:23:22 pm by kawal »
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #88 on: March 01, 2020, 02:11:11 pm »
Quote
TP213 +0.4V
And TP214 is still -2.3V?

Then check voltages on U236 as I asked before. Something is wrong there.
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #89 on: March 01, 2020, 02:32:50 pm »
Quote
TP213 +0.4V
And TP214 is still -2.3V?

Then check voltages on U236 as I asked before. Something is wrong there.

U236 in voltage mode

pin 2 0V
pin 3 -0.4V (depends on current ranage)
Pin 6 -13V (output i guess)
pin 4 -15V
pin 7 +15V


U236 in current mode
 pin 2 0V
pin 3  uVs
Pin 6 -2.3 (output i guess)
pin 4 -15V
pin 7 +15V

In Voltage mode the unit is trying to limit the output as the current its reading is wrong. So based on the data the analog section must correct the situation and the result -240V on output.
In current mode the current measured by the unit  is below the settings  ( I give it a 1mA setting )  and the meter shows less than that with output open.
In current mode the output reaches compliance voltage settings.   I can change the compliance voltage setting and the output follows the setting exactly. 
« Last Edit: March 01, 2020, 02:40:53 pm by kawal »
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #90 on: March 01, 2020, 02:42:11 pm »
Quote
OK  in Voltage mode :
   TP213 in voltage mode is -0.76V  in lower current ranges and -0.4V in higher current ranges
   TP214 is -2.3V

OK, it's starting to make sense...

-0.76V and -0.4V at TP213 are both bad.

TP213 is the output of control inputs mixer.
TP205 (voltage POS), TP206 (voltage NEG), TP207 (current POS), TP208(current NEG) are respective control inputs to that summing thing.

What are the voltages on TP205, TP206, TP207,  TP208 (in voltage mode)
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #91 on: March 01, 2020, 03:51:19 pm »
TP205 9.4V
TP206 4.4V
TP207 -3.8V
TP208 -3.8V
Setting is volts 10.0000V and compileance current is set to  1.05000mA . Reading  Amps +0.75753mA. in lower ranges the meter shows overflow for current.
Reading volts Overflow on
Tp231 is -7.59V
Tp230 is +13.9V
Pin 2 of U227 +3V
Pin 3 of U227 -4.5V
Pin 6 of U227 =TP231=-7.59V (measures 0 ohms between these btw.)

« Last Edit: March 01, 2020, 04:04:14 pm by kawal »
 

Offline openloop

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #92 on: March 01, 2020, 04:34:45 pm »
Quote
TP205 9.4V
TP206 4.4V
TP207 -3.8V
TP208 -3.8V

Assuming these values are correct.  ;)
And in voltage mode, right?

Those readings imply that control signal (and output that follows it) is dragged down by the driver that controls current (negative voltages at TP207 and TP208). Surprisingly enough.
(unless you mixed up polarity again, then the goose chase will continue)

Just to make sure,
What are the voltages at TP209, TP210,TP211 and TP212?
(these are outputs of buffers for TP205, TP206, TP207,  TP208 respectively. Take measurements in the same mode that you had when you've got readings above (so I'm not comparing apples to oranges).

See how many test points they've put in this area? The place is trouble.
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #93 on: March 01, 2020, 05:38:27 pm »
Openloop
I am grateful for the help. and looking at the points you recommended definitely is  making this thing more understandable.
The troubleshooting will need to wait as I have a business trip starting today.
Will be back to it in 2 weeks.  I think thats why I was rushing it and damaged the dacs (I think i did as they measured good before.) Both dacs are running burning hot which I think it is not normal. 

I will be back in 2 weeks to continue this story.

Thanks again for your help
I need to figure out the current measurement section as next priority, That side is definitely sick,.
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #94 on: March 04, 2020, 08:07:58 pm »
A user made me aware that there could be a massive failure of the current sense resistor  switching circuit. The output would in that case show current as there are back to back protection zener diodes across the current sense section.  So if none of the  resistor sections are on,  there would be a about 5V drop across the Current sense Low and Current Sens Hi .  This seems to be what I am experiencing. This would be every bad.  :(
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #95 on: March 05, 2020, 01:58:44 pm »
I made the official request to Tektronix  to release the drawings for the 2400 SMU .  This design must be at least 25 years old by now and I doubt that they would still want to protect that design.  I think the issue is that the current model is the same model number as the one sold in the 1900s and 2000s but the design must be different since there is a lot of obsolete parts in the older 2400.
Lets see what happens .
 
The following users thanked this post: Zucca

Offline Zucca

  • Supporter
  • ****
  • Posts: 3278
  • Country: it
  • EE meid in Itali
Re: Keithley 2400 SourceMeter review and teardown
« Reply #96 on: March 05, 2020, 02:56:29 pm »
Thanks kawal.
+1

#freetheK2400 !

TTi is giving the schematics for old devices to people who ask nicely, I do not know why not all the other are doing the same.
Can't know what you don't love. St. Augustine
Can't love what you don't know. Zucca
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #97 on: March 05, 2020, 06:42:09 pm »
I asked nicely but Tektronix  did not  like my request.  Gave me the link to the manual I said I already have.
Here is the reply :

   Techsupport
    Thu, 03/05/20 5:09 pm   
 
    Service manual is what we offer https://www.tek.com/keithley-source-measure-units/keithley-smu-2400-series-sourcemeter-manual. Schematics are not available to general public.

 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #98 on: March 09, 2020, 07:10:29 pm »
Got a new reply from Tektronix but the story is the same.  No we cannot give you schematics.

KI-Support-Apps <ki-support-apps@keithley.com>
Attachments
9:47 AM (5 hours ago)
to me

Hello,

 

Unfortunately, Tektronix and Keithley no longer provide schematics for our products, as they are considered IP that is unavailable to customers.

This policy was instituted before the release of the Model 2400.

The best resource for self-repair is the manual. Otherwise, we would recommend working with our Service Department to schedule a repair appointment.

 

Best Regards,

Jake Austin

Applications Engineer
 

Offline kawal

  • Regular Contributor
  • *
  • Posts: 188
  • Country: us
Re: Keithley 2400 SourceMeter review and teardown
« Reply #99 on: March 14, 2020, 04:50:59 am »
Made major progress after coming back from business trip.
I changed the current DAC and the meter started working . After that changed the overheating Voltage DAC and now the unit seems functional in Voltage ranges.  I  made a rough calibration as the display was out quite a bit out after DAC changes. When all done I will send the unit out for calibration. Dacs are warm but not anything near the temperatures as the old ones.

Next issue is the output always on issue .  I found the issue with that as well. This issue is caused by damaged U253 74HC4094. The output pin 5 controls the output solenoid and it always is high no matter what. Grounding the pin releases the ouput relay.
I checked the U256 which switches current ranges and that works  perfectly.  U255 switches voltage dac and gains for the voltage input amp and seems fine. U254/U252 did not measure as I am not 100% how to check that.

Will need to by some shift registers as next purchase.


« Last Edit: March 14, 2020, 05:15:49 am by kawal »
 
The following users thanked this post: TiN


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf