Author Topic: HP 8903B voltage drop at higher frequencies  (Read 4864 times)

0 Members and 1 Guest are viewing this topic.

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
HP 8903B voltage drop at higher frequencies
« on: October 30, 2015, 07:33:48 am »
Recently bought a HP 8903b everything seems to check out but the power is dropping out at higher frequencies. When I set it to 1 KHz and 1 volt it reads 1000.7 Hz and 1.000 Volts which seems good to me but @ 100 KHz it reads 0.387 Volts, so I'm wondering if the signal generator transistor or power supply has something wrong with it that's causing it to lose power at higher frequencies? It begins to gradually lose power around 6 KHz. Thoughts?
 

Offline macboy

  • Super Contributor
  • ***
  • Posts: 2011
  • Country: ca
Re: HP 8903B voltage drop at higher frequencies
« Reply #1 on: October 30, 2015, 01:16:50 pm »
Have you verified the output voltage with another voltmeter? If not then you don't even know if the fault is in the oscillator/output section or the input/measurement section. When measuring the output, keep in mind the bandwidth of the meter you are using. Most handheld meters are not flat to 100 kHz. You will need to use a good quality bench meter or an oscilloscope at that frequency.
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #2 on: October 30, 2015, 11:33:55 pm »
Here are the readings,I used an HP 54200A Digitizing Oscilloscope and a Tenma 72-7735 digital MM.  I tied the oscilloscope in to the HP 8903 with a BNC T and took reading with a digital MM.

   HP 8903b set to                      HP 8903 Readings                 Oscilloscope Readings                 MM Readings
   20 Hz                                      19.99Hz   0.997v                   20Hz    2.75v                               0.991v
   100Hz                                     99.98Hz   1.000v                   100Hz  2.79v                               0.993v
   500HZ                                    500.29Hz  1.000v                  498.8-501.3Hz 2.79v                  0.990v
   1KHz                                      1001.2Hz  1.000v                  998-1004Hz  2.73v                      0.980v
   5KHz                                      4994.3Hz  0.998v                  4975-5013Hz  2.73v                    0.779v
   10KHz                                    10002Hz   0.993v                  9980-10020Hz  2.73v                  0.519v
   20KHz                                    20007Hz   0.973v                  19960-20040Hz  2.61v                0.253v
   50KHz                                    50018Hz   0.821v                  49750-50250Hz  2.33v                0.014v
   100KHz                                  99985HZ  0.363v                   99800-100200Hz  1.98v             0v   
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #3 on: October 31, 2015, 12:50:26 am »
Check A3C2 & A3C3.  These are 20 µF 50 V capacitors on the 2nd card in the card cage.
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #4 on: October 31, 2015, 01:41:05 am »
I think I found the caps but, does anyone have a circuit diagram from the HP 8903? I found the user manual and the service manual but, I haven't been able to find circuit diagrams.

  To test them I would have to remove them from the board correct? Should I just go ahead and change them? When I do work on older amplifiers I typically change out the electrolytic capacitors if they're over 20-25 years. If I do need to replace them is there any one capacitor that's better than another? or could I just use Nichicon Fine gold caps? And is there any other components in these that are known to go with with age?

                 Thanks
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #5 on: October 31, 2015, 02:27:07 am »
There are two service manuals, and a calibration manual.  They are available at the Keysight website, but you'll need to create an account.

Caps with values of electrolytes can usually be tested in-circuit with a good LCR meter (if it uses a bridge circuit).  If they are so bad that they are pulling down the circuit, then even a cheap ESR meter should work.  I would not trust the capacitance function of a DMM for an in-circuit test whatsoever.

I'd replace those with 22 µF mil spec tantalum capacitors (and they are not cheap). They are on the ± 15 Vdc input lines of the PCB.
OEM: http://www.newark.com/vishay-sprague/te1305/aluminum-electrolytic-capacitor/dp/18F531
What I'd use: http://www.digikey.com/product-detail/en/M39003%2F01-6234/1012-1075-MIL/3190972
35 Vdc cap should work on the 15 Vdc rails: http://www.digikey.com/product-detail/en/M39003%2F01-6111/1012-1068-MIL/3190965
« Last Edit: October 31, 2015, 02:55:08 am by dacman »
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #6 on: October 31, 2015, 03:24:03 am »
You're not kidding those are pricey caps, I think I'm going to hold off on that upgrade for the time being and just go with the cheaper ones and see if it corrects the problem. I'll let you know what happens, Thanks for the Help!  :-+
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #7 on: November 09, 2015, 02:04:38 am »
Just put in the new capacitors, the voltage levels have slightly increased but, I also realized I'd made a stupid mistake when taking my reading. I had the 80 KHz filter on :p with it off I get 0.78 v on the HP8903 and 2.05v on the oscilloscope @100 KHz, 1 volt. And the distortion reading at 100 KHz is 0.0358 and at 1KHz is 0.0264 that's with no filters on. Does that sounds about right or is something off?
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #8 on: November 09, 2015, 03:38:36 am »
Voltmeter reading at 100 kHz is low and distortion at 1 kHz is high (100 kHz distortion is OK).  What readings do you get with the output set to 6V?

Go to section 5 of the calibration manual.  Most of the adjustments are for the meter.  There are several flatness adjustments for the attenuator.

For adjustment 6, the output can be connected to the input and then it can be adjusted for the lowest distortion (-90 dB @ 1 kHz which is about 0.0032 %).  The spec at 1 kHz is -80 dB (or 0.0100%).

The only other common issue with these instruments that I know of is the Float Chassis switches sometimes need to be cleaned (but not very often).  They would need to be lubricated after cleaning (or just lubricated).
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #9 on: November 09, 2015, 05:34:00 am »
With all the filters off and it set to 1KHz @ 6 volts the distortion reads 0.0059, with 100KHz @ 6 volts the distortion reads 0.0880
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #10 on: November 10, 2015, 02:07:55 am »
With all the filters off and it set to 1KHz @ 6 volts the distortion reads 0.0059, with 100KHz @ 6 volts the distortion reads 0.0880

Slightly OOT at 100 kHz.  Unit needs tweaked.
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #11 on: November 10, 2015, 03:13:27 am »
Do you think something needs to be replaced or does it just need to be re-calibrated? I tried starting at the beginning of the trouble shooting in the service manual and it instructs you to:

 1. On the Audio Analyzer, key in 41.0 SPCL to initialize the instrument. Set the INPUT and OUTPUT switches
to ground. Set IMPEDANCE to 600R. Key in AMPTD 3 V. Connect the HIGH OUTPUT to the HIGH INPUT
through a tee adapter.
2. Connect a high-impedance, ac coupled oscilloscope to the tee at the HIGH INPUT. Use the oscilloscope’s
vertical gain controls to adjust the display amplitude for 5 divisions peak-to-peak deflection of the sinusoidal
signal. (The period of the signal should be 1 ms.)
3. Connect the oscilloscope to A2TP5 (SING END). Key in the Special functions listed in Table 8B-4. For each
setting, increase the oscilloscope’s vertical gain by the amount indicated. The peak-to-peak amplitude, displayed
on the oscilloscope, should be within the limits indicated. If faulty, see Service Sheet 1 and check the High Input
Attenuator.

 Now I'm new to using an oscilloscope so I'm pretty sure I'm doing it wrong, I set it to auto scale and to a/c coupling and when I punch in V ampl measure channel I get 8.18 V so I tried V max and got 4.14v but according to the chart it's suppose to be between 4.9 and 5.1 amplitude limits (divisions peak to peak) 
So if I'm making a stupid mistake in my settings please correct me.
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #12 on: November 10, 2015, 04:08:32 am »
I'd do the calibration first.  I would not know whether or not a repair was needed until after the calibration was attempted.  Your unit isn't that bad, it's just out-of-tolerance.

You need to take your oscilloscope out of auto mode.  They want you to adjust for a 5 division peak to peak reference in step 2 using the oscilloscope's amplitude vernier (and V/DIV control), then connect to A2TP5 and check for the same 5 divisions in step 3.  Later, when instructed to increase the oscilloscope's vertical gain by X10, if it were set to, for example, 1V/DIV, you would then set it to 100mV/DIV leaving the vernier alone.
« Last Edit: November 10, 2015, 04:13:23 am by dacman »
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #13 on: November 10, 2015, 07:40:32 am »
When I attempt the DC offset adjustment in the calibration section the lowest it will go is -62.5 mVdc but the manual wants it to be 0 +/- 0.1 mVdc.
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #14 on: November 11, 2015, 12:45:45 am »
Check if A6TP4 is varying (referenced to A6TP1 or GND).  If you have a BNC short, then place it on the OUTPUT LOW connector and see what happens.
 

Offline Simon Spiers

  • Regular Contributor
  • *
  • Posts: 76
  • Country: gb
    • My Youtube channel
Re: HP 8903B voltage drop at higher frequencies
« Reply #15 on: November 12, 2015, 08:54:45 pm »
Make sure your termination is correct.
I would sweep the output through a scope of adequate bandwidth and see if you can see any rollout.

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #16 on: November 12, 2015, 10:13:42 pm »
Checked A6TP4 refrenced to ground not shorted reading was 1,002 Hz @ 2.631 Vrms Shorted gave me the same reading (had it set to 3 volts)
« Last Edit: November 12, 2015, 10:44:19 pm by thinker1920 »
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #17 on: November 12, 2015, 10:59:40 pm »
When I attempt the DC offset adjustment in the calibration section the lowest it will go is -62.5 mVdc but the manual wants it to be 0 +/- 0.1 mVdc.

If this is Adjustment 9, and in step 3 you are measuring -62.5 mVdc at the OUTPUT High jack, then this could be caused by the Float Chassis switch on the OUTPUT side.  Place a BNC shorting cap on the OUTPUT Low jack.  This is equivalent to setting the Float Chassis switch to Chassis.  (I know the manual uses ground vice chassis, but the symbol on the front is the chassis symbol.)  If the output can then be adjusted to 0 mVdc, then the Float Chassis switch needs cleaned (I'd use LPS-1 on the switch).

http://www.amazon.com/s/ref=nb_sb_noss_2/181-8212496-9497506?url=search-alias%3Daps&field-keywords=BNC+Shorting+cap&rh=i%3Aaps%2Ck%3ABNC+Shorting+cap

I'd want the 3rd one in the list above (the Amphenol).

Also, The 8903B came with four of the below adapters (although they were gray).

http://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=Pomona+3430&rh=i%3Aaps%2Ck%3APomona+3430
« Last Edit: November 12, 2015, 11:50:50 pm by dacman »
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #18 on: November 13, 2015, 12:21:41 am »
I shorted the Low output and it works like a charm, I took apart the front and sprayed deoxit in the float switch and flipped it back and forth a bunch and put it back together and it's still not working :(, again if I take a piece of wire and short the low it works great I get 0.985 volts at 100 KHz and was able to adjust my DC offset to 0.1 mV. Thoughts? I was pretty thorough with the cleaning.
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #19 on: November 13, 2015, 12:44:22 am »
I've seen LPS PreSolve work when DeoxIT would not.

Try some contact cleaner or IPA.  (You would need to lubricate the switch after using one of these, else it could be destroyed for lack of lubrication.  I use LPS-1.)

If you can get to the contacts, try an ink eraser.

If you want to know what I would have tried at the start, I would have drenched it in LPS-1 (and if this did not work, I would have then tried the above).

In the meantime, I recommend getting a BNC Shoring Cap, and just using it (a minor issue).
« Last Edit: November 13, 2015, 12:57:42 am by dacman »
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #20 on: November 13, 2015, 12:51:42 am »
Thank you for seeing me through this!!! I really do appreciate it :)  :-+
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #21 on: November 13, 2015, 01:19:58 am »
The value you were getting, -62.5 mVdc, is pretty bad, almost as if the switch is open, and if it were adjusted down to that, then maybe the switch is open.  Could you check the continuity of the switch connections to see if it is actually connected to the motherboard?  Could you measure the resistance across the switch?
 

Offline thinker1920

  • Contributor
  • Posts: 19
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #22 on: November 13, 2015, 02:50:09 am »
It was the float switch and at the same time it wasn't :P I dissembled the actual switch and cleaned the internal contacts which were black, I used a Q tip with deoxit, I reassembled it and measured it with my DMM, and it checked out. I reassembled and turned it on and got the same readings :p So I thought maybe the wire that connected it to the mother board was bad so I dove into it again and disconnected the molex type connector and used my DMM and the wire was good. So I then took off the bottom of the HP8903b and thought I would follow the traces, and it was right where the connector was solder to the board, a trace that looked like it had more or less vaporized, I soldered in a small piece of wire where the trace use to be and voila! All my readings are within tolerance, DC offset is now 0 mVdc, set at 1KHz @ 1 volt reads 1000.9 Hz @ 1.009 volts with .0024 distortion (with 80 KHz filter), set to 100KHz @ 1 volt reads 99,891 Hz @ 1.004 volts with .0214 distortion (no filter). Are those acceptable numbers? 
 

Offline dacman

  • Frequent Contributor
  • **
  • Posts: 404
  • Country: us
Re: HP 8903B voltage drop at higher frequencies
« Reply #23 on: November 13, 2015, 03:53:19 am »
All those numbers look fine.  -92 dB residual at 1 kHz is very good.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf