Author Topic: Keithley 2000 (Read 26505 times)

EMUD · « **on:** March 15, 2015, 09:01:44 pm »

Hi,
I just try to understand the A/D converter of the K2000. It is a multi slope converter (it converts a current to a pulse length).
I would like to make some calculations about this converter and need some help.

If possible check your Reference voltage, the + 14V and the -14V (these voltages are regulated by a double op amp and a transistor each) and post them here.

Thanks a lot for your help!

Thanks
EMUD

EMUD · « **Reply #1 on:** March 16, 2015, 12:14:50 pm »

Hello,
I have attached a block diagram of the A/D conversion unit. I think that this unit measures the voltages selected by U163
with comparism to the 4 voltages (0 volts (GND), 0,09811V, 1,0320V, 6,95V).

There are some A/D related tests:

Test 100.1 checks the charge - discharge circuit of the integrator formed by U137 and the components around. Measurement input is GND.
Test 100.2 checks the differences between some measurement results.
Test 101.1 does the same as 100.1 , but stores the result.
Test 101.2 does the same as 101.1 , but measurement input is Vref (6,95V nom.)
Test 101.3 checks the difference between 101.1 and 101.2 .

Best regards

EMUD

TiN · « **Reply #2 on:** March 16, 2015, 12:21:38 pm »

So what exactly you want us to do? I have broken 2000 which I can measure for you.
Just +14V/-14V output from opamps?

EMUD · « **Reply #3 on:** March 16, 2015, 12:50:34 pm »

Hello TiN,

do not measure the +14/-14 at the op, since it is a simplified diagram, there are 2 transistors in the real world to
boost the current.
The easiest way to contact this is the resistornetwork TF-245 on position R271.
Please measure with respect to analog ground pin 1 (should be +14V) and pin 3 (should be -14V) and the
reference voltage at R315 to ground (should be +6,95V).
These voltages could be measured with the K2000 itself, you only need the positive input and the volts range.

Thank you in advance!

EMUD

P.S. The schematic and the repair manual are downloadable free in the i-net ! If I can help you with your 2000 ...

TiN · « **Reply #4 on:** March 16, 2015, 01:52:22 pm »

We know, but schematics is reversed by bbs.38hot.net member, not from vendor

Now, why it was important?

ElektroQuark · « **Reply #5 on:** March 16, 2015, 01:57:50 pm »

Hi EMUD,

I'm trying to repair a K2000.
Here is the thread: https://www.eevblog.com/forum/testgear/restoring-and-upgrading-rom-of-a-keithley-2000/

Any help/ideas are appreciated. Thank you.

EMUD · « **Reply #6 on:** March 16, 2015, 02:15:48 pm »

Hi TiN,

thanks for the values! I need this values to verify the values in the schematic. So, if I could collect
the values of some K2000 units we will see the tolerances and make it easier to hunt for problems.
I have seen that the +14/-14 should be double the height than the Vref. In the reversed diagram
there are some voltages given wich could lead to the wrong results in hunting for errors.

Your results : Vref 6,8470373 -V14 13,6860304 +V14 13,6983588

In the pics you have posted I see that you are quite often repairing Keithleys. Many of them use
the resistor network TF-245 . What is your opinion, does it fail often?

Thanks a lot! I will post the results here.

EMUD

TiN · « **Reply #7 on:** March 16, 2015, 04:34:00 pm »

Not so many, I mostly have 2001, which does not have TF-245 network. 2400 and 2510 have it, since they leverage almost same ADC block as 2000.
I got 2000 while ago, but had no much progress on it, so far. I saw one failed unit with TF-245 on chineese 38hot, but that's it.

I had thread about my 2000 here.

z01z · « **Reply #8 on:** March 17, 2015, 07:18:05 am »

Another TF-245 failure is described here. There's also info on what's inside.

TiN, you'll have to append a backslash to the topic link to make it work.

EMUD · « **Reply #9 on:** March 17, 2015, 07:36:44 am »

Hello z01z,

yes, I have read this before and put the reversed schematic and the values of the costas TF-245 together in my block diagram.
I have calculated some of the values and they show that the costas TF245 schematic and the values are ok.

Now, if we have the values of the resistors inside the network and its schematic, we only need the tolerance to test it. Calculating with the values I have got from three K2000 it has to be a high precision network, I think with 0.1 or 0.05% . So I could state that my TF-245 is gone, one of the 10K resistors has 10.5K ....

The network is available at Keithley service here in Germany for about 16€ + VAT + p&p . I have ordered and looking forward to get it (about 4 weeks).

Many thanks to all for the information

EMUD

klaus11 · « **Reply #10 on:** March 17, 2015, 07:43:56 am »

Quote from: EMUD on March 17, 2015, 07:36:44 am

Hello z01z,

yes, I have read this before and put the reversed schematic and the values of the costas TF-245 together in my block diagram.
I have calculated some of the values and they show that the costas TF245 schematic and the values are ok.

Now, if we have the values of the resistors inside the network and its schematic, we only need the tolerance to test it. Calculating with the values I have got from three K2000 it has to be a high precision network, I think with 0.1 or 0.05% . So I could state that my TF-245 is gone, one of the 10K resistors has 10.5K ....

The network is available at Keithley service here in Germany for about 16€ + VAT + p&p . I have ordered and looking forward to get it (about 4 weeks).

Many thanks to all for the information

EMUD

Did you also sell the transistor snj132 in Keithley Germany?

EMUD · « **Reply #11 on:** March 17, 2015, 08:07:03 am »

Hello Klaus11,

I have sent you a PN with my contact to Keithley Germany.

EMUD

klaus11 · « **Reply #12 on:** March 17, 2015, 08:44:01 am »

Quote from: EMUD on March 16, 2015, 12:50:34 pm

Hello TiN,

do not measure the +14/-14 at the op, since it is a simplified diagram, there are 2 transistors in the real world to
boost the current.
The easiest way to contact this is the resistornetwork TF-245 on position R271.
Please measure with respect to analog ground pin 1 (should be +14V) and pin 3 (should be -14V) and the
reference voltage at R315 to ground (should be +6,95V).
These voltages could be measured with the K2000 itself, you only need the positive input and the volts range.

Thank you in advance!

EMUD

P.S. The schematic and the repair manual are downloadable free in the i-net ! If I can help you with your 2000 ...

TF245 Pin 1 +14,03 Pin 2 -14,02 to Pin 6 AGND

EMUD · « **Reply #13 on:** March 17, 2015, 08:59:43 am »

Hi Klaus11,
I have calculated your Vref : 7.015V Is this ok ?

Thanks

EMUD

picburner · « **Reply #14 on:** March 17, 2015, 02:13:39 pm »

Hi Emud,

The resistor network TF-245 appears to have a failure rate of 43.2%.
see http://bbs.38hot.net/thread-94614-2-1.html (better to use google translate...)
I have a Keithley mod. 2306 (2 TF-245), one mod. 2010 and two mod. 2000.
The 2306 has the two TF-245 perfects as well as one of the two mod. 2000.
In the 2010 the TF-245 has one resistor open and the other 2000 has 4 resistor out of tolerance.
This seems confirm the statistic.
Could you give me also your contact at Keithley Germany?
I need 2 TF-245, thanks.

Regards
Giampiero

EMUD · « **Reply #15 on:** March 17, 2015, 03:15:21 pm »

Hi Picburner,

what do you think, are there special resistors affected or is it random? (poor quality??). Have you any information about the tolerance (I think min. 0.1%)?

What is your reference value to find the resistors faulty?

Thanks a lot

EMUD

picburner · « **Reply #16 on:** March 17, 2015, 04:11:11 pm »

Hi Emud,

I don't know the exact tolerance of the 12 resistors in the TF-245 but considering the type of instrument I agree with you for a 0.1%.
In my case I found a 20K resistor become fully open, 90K become 95K, 30,176K -> 31,365K, 1K -> 1,15K etc.
I found the internal TF-245 schematic on bbs.38hot.net.

Best regards.
Giampiero

EMUD · « **Reply #17 on:** March 17, 2015, 04:20:10 pm »

Hi Picburner,

that is my issue, a 10K has become 10.5K, so the reference is not correctly amplified to the 14 / -14V rails. (6.91V ref, 13,5V , - 13,2V)
But I think the values in the schematic are not correct. There is a schematic with other values in the net, I think this one is ok. I have attached
the .pdf .

Thanks

EMUD

picburner · « **Reply #18 on:** March 17, 2015, 04:32:57 pm »

Yes, the resistors arrangement is the same but some values don't agree....
What will be right?

Regards
Giampiero

EMUD · « **Reply #19 on:** March 17, 2015, 05:35:31 pm »

In my opinion this is these are the correct values, I think they are measured on a new TF-245, but I am not sure.

Emud

EMUD · « **Reply #20 on:** March 19, 2015, 04:49:55 am »

Here are the result of my TF-245 network, desoldered :

Pin 1 - 16 1.0013K
Pin 16 - 2 9.037K
Pin 2 - 15 10.467K
Pin 15 - 3 20.07K
Pin 3 - 14 15.04K
Pin 14 - 4 30.88K
Pin 4 - 13 30.89K
Pin 5 - 11 30.42K
Pin 6 - 9 1.0006K
Pin 7 - 9 9.418K
Pin 7 - 8 92.24K
Pin 10 - 12 27.56K

I think its gone ...

Now I am waiting for the new network, I will post the values of the new one (in about 4 weeks ...).

EMUD

ElektroQuark · « **Reply #21 on:** March 19, 2015, 05:48:26 am »

Those are the values of mine, for another reference:

PIN OHMS PIN

1 999.58 16
2 8997.6 16
2 9998.4 15
3 19K994 15
3 14K995 14
4 30K171 14
4 29K698 13
5 30K163 11
6 999.65 9
7 8997.5 9
7 89K974 8
10 27K412 12

picburner · « **Reply #22 on:** March 19, 2015, 07:55:46 am »

Hi EdoNork,

yours measured values seem to fall within the 0.05% tolerance according to
the internal schematic posted by me earlier.
Stressing it as little as possible with the heat because it seems perfect.
The Emud's TF-245 is gone but not so worse than 2 of mine.

Regards
Giampiero

TiN · « **Reply #23 on:** March 19, 2015, 04:20:32 pm »

Measured my 2000's TF-245.
Result: http://dev.xdevs.com/issues/1178
Matches schematics nicely.

picburner · « **Reply #24 on:** March 20, 2015, 07:03:27 am »

The argument TF-245 seems to excite a lot so I put some photos here.
the first photo shows what's inside TF-245,
the second and third photo shows the circuit that replaces the TF-245
in the new DMM pcb revision (2000 and 2015).
The fourth photo shows the pcb's layout of the old mod. 2000.
I found this on the forum bbs.38hot.net always, as seen from the watermark on the pictures ....

EMUD · « **Reply #25 on:** March 20, 2015, 08:00:37 am »

Hi,
a view over all the values seemed to give the following information:

1. R271 (TF-245) is a thin film coustom resistor network with 0.1% or better tolerance.
2. The values given in the reversed schematic are meeting the measured TF-245 s.
3. For some resistors (i.e. the 1K,9K and 90K) the pecise value is not the problem,but the ratio.
4. For some resistors the value has to be precise, i.e. the charging Rs for the integrator.

OK,
now it should be no problem to check the own TF-245.

Thanks

EMUD

EMUD · « **Reply #26 on:** April 02, 2015, 07:47:57 am »

Hello,
the new TF-245 has arrived. Since I have got 4 arrays I have checked the values.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11

0.9999K 8.9999K 10.0001K 20.000K 14.9995K 30.173K 29.708K 30.178K 27.421K 1.0003K 9.0040K

1.0002K 8.9989K 9.9985K 20.000K 15.0035K 30.168K 29.697K 30.166K 27.415K 1.0000K 8,9990K

1.0002K 8.9999K 9.9995K 19.996K 14.9973K 30.173K 29.700K 30.176K 27.416K 1.0001K 8.9993K

0.9997K 8.9960K 10.0030K 19.990K 15.0012K 30.176K 29.704K 30.174K 27.417K 1.0006K 9.0007K

R12

90.001K

89.979K

89.982K

89.990K

I have replaced my faulty TF-245 and all the problems are gone.

But I have discovered a strange behaviour with the self test:

Software A02

When selecting ACV or FREQ and starting selftest, all tests failed! Starting self test after starting the unit there is no problem with the selftest !!

Has anyone seen this, too ?

Best regards and a nice easter weekend

EMUD

EEVblog · « **Reply #27 on:** April 03, 2015, 01:23:50 pm »

miguelvp · « **Reply #28 on:** April 05, 2015, 03:08:47 am »

somewhere in page two of this thread something went really wrong, I just get a blank page and the only way to read it is by pressing reply and read the messages bellow

picburner · « **Reply #29 on:** April 23, 2015, 04:29:17 pm »

To the Administrator:
can you repair the page 2 of this thread?
It has become invisible, thanks.

Robert · « **Reply #30 on:** June 18, 2021, 01:45:15 pm »

Hello Everybody,

Sorry for resurrecting this quiet old topic.

I have a Keithley 2000-20 with quiet strange behaviour.
Most times it works normal but I don't have another 6.5 digit DMM (or better) to check accurecy of the last digits, so it could be slightly off.

Sometimes it is measuring complete bullshit and selftest 101.2 and following will also fail than. This is absolute random and sometimes it fixes itself, somtimes after powering on and off or switching ranges.
I already replaced the relais (and the tweeter by the way) but this haven't made any difference.

After several weeks of short trubelshooting sessions (the error is wenting away when you hunt it) I'm pretty sure that my TF-245 is gone.
The 27.42K resistor between pin 10 and 12 goes open when I heat it up.
I can measure it in circuit when powered off (Q127 and Q131 are non conducting by default). In non failure state the value is 27.468K (in circuit) not bang on but close. When I heat it up the resistance rises slowly until it suddely changes to infinite. This corresponds to the voltage measured on pin 10 when stopping after test 101.2, if it passes, I can measure 6.507 V as result of the voltage divider of R300 and the resistor inside the TF-245. When the tests 101.2 fails the voltage of pin 10 is at 7.015 V, exactly the reference voltage which indicates that no current flows through R300.

Maybe this description helps other people with similar problems.

Could someone please help me out with a source of a spare part? I'm located in Germany and haven't contacted Keithley/Tektronix yet.

Best regards and a hot weekend
Robert


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Keithley 2000 (Read 26505 times)

Share me