Forums :: NEW! Web Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Nov Mon 29, 2021 9:59 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 10 posts ] 
Author Message
 Post subject: HP 8555A ADVICE NEEDED
PostPosted: Aug Sun 29, 2021 3:56 pm 
New Member

Joined: Jan Mon 20, 2020 3:24 pm
Posts: 6
Hello I am repairing the HP spectrum analyzer type 141T and the unit HP8555A.
There were many defects in it but these have all been resolved. One problem remains and that is about 12 dB too low a reproduction.

After several measurements, I think this error is in the 50 MHz AMPL ASSY A2A5 (HP 08555-60060).

The input of this unit is a JFET (Q1) and 4 transistors (q2-q5) type 2N7179.
In the service documentation (SERVICE SHEET 5) a number of voltages are specified with a tolerance of 0.1 V. The measured voltages at the JFET on the Source is 2V (must be 2.8V) and on the Drain 13.00V (must be 10.06V .

Also all other voltages stated on the transistors Q2-Q5 are about 2 Volt higher on the Emtter and Base. The voltage on the Collector is good (19.9V).

What is striking is that the JFET is very warm > 60 degrees.

Unfortunately no AC gain value has been specified. I measure a gain on the first unit Q1-Q2 of 30 dB. The +15 dB and +5 dB circuit work well.

Suspicious is of course the JFET but it is difficult to buy and very expensive. In order to have a better look at this 50 MHz Amply A2A5 and possibly replace parts, you should remove the PCB.

The service documentation SERVICE SHEET 18 describes the REMOVEL AND REPLACEMENT PROCEDURE of the THIRD CONVERTER A2 . In order to remove the entire module housing and board A2A5, the wires at the bottom of the housing must be soldered loose.

What is not clear to me is whether when removing the housing of A2A5 the connections A2C8, A2C10, A2C11, A2C12, A2C8 and A2C13 on the board are solder points or whether these are pin connectors.

When I have removed the screws from the A2 board and removed the connection 50MHz input and the soldering to J4 and J2 I can not get the PCB out of the housing. The PCB remains attached to connections A2C8 to A2C13.

So my questions are:
1st Does anyone know if the JFET q1 gets that hot
2nd Are the connections A2C8 to A2C13 solder joints or connectors.

Thank you in advance for your response, best regards,

L.R.


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Aug Mon 30, 2021 1:01 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 6666
Location: Montvale NJ, 07645
You might get better responses posting on this forum. Lots of really smart HP guys there:

https://groups.io/g/HP-Agilent-Keysight ... ent/topics


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Aug Mon 30, 2021 3:41 pm 
Member

Joined: Jul Wed 22, 2009 3:07 pm
Posts: 937
Hi LR,

With the nominal voltages the dissipation of the J-FET is about 120mW, so it is conceivable that it gets that hot.
J-FETs have quite a variation in their parameters (pinch-off voltage, Idss), so 2V at the source instead of 2.8V seems acceptable to me.

A quick simulation of the input stage (Q1, Q2) shows that the gain is about 30dB, it can be adjusted by the R4 pot (I-LOW) from about 27dB to 36dB. (This is by using a 50 Ohm generator, a J310 FET, and a MPSH10 bjt in the simulation.)
The stage needs to be tuned up with C2 (INPUT MATCH) to 50MHz.
Changing the gain by R4 also mis-tunes the input match, so you need to touch it up by adjusting C2 after changing R4.

EDIT:
Also do check the first mixer!
A blown mixer could also show reduced sensitivity in the SA.
Checking the mixer is detailed on page 8-16 and the page before in the OP/Service manual.

Hope this helps, Peter


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Aug Mon 30, 2021 9:05 pm 
New Member

Joined: Jan Mon 20, 2020 3:24 pm
Posts: 6
Hello Peter,

Thanks for your comment. As for the deviations of the DC voltages, they should only deviate 0.1V according to the service documentation.

The gain I measured from C1 to base Q3 is 30 dB with R4 in the middle position. For the time being, I exclude other causes of the too low (12dB) display. According to the service documentation I have to transmit with -40 dB on c1 and then the prompt on the screen should be -30 dB. In this case, the indication -12 dB is too low. According to the service documentation, the output of the 50 MHz amplifier should have -33 dB. This is also 12 dB too low in my case. If I transmit on J2 with -33 dB, then the indication is correct, so -30 dB on the screen. So far everything points to the fault being in the 50 MHz amplifier.
A number of parts on this PCB cannot be measured accurately by connecting parts in parallel.

To be able to examine parts better, I would have to remove the 50 MHz PCB. And that's what my next question is about. The service documentation describes how to remove the housing including the 50 MHz amplifier to access the underlying 500 MHz amplifier. You have to remove a number of screws and solder the wires to A2C8 to A2 C13.

But how do you remove the 50 MHz amplifier from its housing? After removing the screws and the wires to J1 and J2, the PCB remains attached to capacitors A2C8 to A2C13.

It is not clear to me whether these points are also solder points or whether these are connectors.

I took some pictures for clarity.

Attachment:
c-oppr.jpg
c-oppr.jpg [ 976.5 KiB | Viewed 635 times ]


Attachment:
draad-c-kl.jpg
draad-c-kl.jpg [ 960 KiB | Viewed 635 times ]


I hope you can still help me with that. Thanks again with kind regards,

loek


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Aug Mon 30, 2021 11:59 pm 
Member

Joined: Jul Wed 22, 2009 3:07 pm
Posts: 937
Hi Loek,

I did see your note about the 0.1V tolerance, but I still believe that your input stage is just fine.
My guess is that selecting FETs for those voltage tolerances were increasingly getting harder for HP, and they eventually relaxed the tolerance requirements. But this is just a guess though.
The fact that the simulated gain results match your measurements, while the source voltage in the simulation with the J310 FET is 1.75V, tells me that the input stage in your amp is basically fine.

I am wondering if those soldered connections shown at the bottom of your first picture are the female connectors where pins connect to those.
See if you can pry the board up gently when the screws are removed. It is also possible that those connections are soldered connections.

Before you take that board out and check individual components on it, I am wondering if you went through the full adjustment procedure of that amp stage.
My suggestion is that you follow the amplifier adjustment procedure as described in the service manual.
What I read is a 37-step procedure, in paragraph 5-30, starting on page 5-17 in the manual.
The amplifier itself has PIN diodes in all the following stages to the input stage, and the overall gain of the 50MHz amp depends on the control signals to those PIN diodes.
That is why I am asking if you went through the whole amp adjustment process as laid out in the manual, and if it failed somewhere, where did it it fail?

I also take that you checked the all the supply voltages.

Regards, Peter


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Aug Tue 31, 2021 12:07 pm 
New Member

Joined: Jan Mon 20, 2020 3:24 pm
Posts: 6
Hallo Peter nogmaals bedankt voor je reactie,

De specificatie staat in service sheet 5 zie de foto below.


Ik heb ook de volledige 37-step procedure gevolgd. Het resultaat bleef hetzelfde dus 12 dB te laag.
In deze procedure komt ook de PIN diode CR2 en CR3 in actie en deze werken goed er is dus een versterking van 5 of 15dB. De invloed van diode CR1 heb ik geƫlimineerd door door punt C9 naar aarde kort te sluiten. De versterking neemt dan met 10 dB toe. Deze 10 dB versterking is dus het regel gebied om afwijkingen in de ingangsmixer te compenseren.

Er moet toch een foute component zitten op het 50 MHz versterker board. Om dit beter te kunnen onderzoeken moet ik dit board kunnen loshalen uit de behuizing. Ik heb geprobeerd zoals jij ook al voorstelde om het board voorzichtig los te halen. Maar ook na stevig trekken komt er geen beweging in. Aan enkele solderingen op het board en ook aan het vele soldeer hars wat op de aansluitingen van A2C8 tot A2C13 zat lijkt het er op dat iemand al eerder een reparatie heeft uitgevoerd. Misschien is er daardoor soldeer op de connector punten gekomen en zit de print daardoor vast.

Ik had gehoopt dat iemand ervaring had met dit board zodat ik zeker weet wat te doen. Het enige wat overblijft is proberen op ik met een tin zuiger de pennen los kan krijgen van de print.

Nogmaals bedankt met vriendelijke groet,

Loek

Hi Peter, thanks again for your comment,

The specification is in service sheet 5, see the photo below.



I also followed the full 37-step procedure. The result remained the same, so 12 dB too low.
In this procedure the PIN diode CR2 and CR3 also come into action and they work well so there is a gain of 5 or 15dB. I eliminated the influence of diode CR1 by short-circuiting point C9 to ground. The gain then increases by 10 dB. This 10 dB gain is therefore the control range to compensate for deviations in the input mixer.

There must be a faulty component on the 50 MHz amplifier board. In order to investigate this better, I need to be able to remove this board from the housing. I tried as you suggested to carefully remove the board. But even after pulling firmly, there is no movement. Some soldering on the board and also the large amount of soldering resin that was on the connections from A2C8 to A2C13, it seems that someone has already carried out a repair. Maybe that's why solder got on the connector points and the PCB is stuck.

I was hoping someone had experience with this board so I know for sure what to do. All that remains is to try to get the pins free from the PCB with a tin plunger.

Thanks again with kind regards,

Loek


Attachments:
spec-tota-kl.jpg
spec-tota-kl.jpg [ 145.71 KiB | Viewed 589 times ]
Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Aug Tue 31, 2021 4:40 pm 
Member

Joined: Jul Wed 22, 2009 3:07 pm
Posts: 937
Hi Loek,

I entered the whole circuit into LTspice.
Here are some points regarding the simulation:

- For the PIN-diodes I used the Rohm RN731 PIN-diode, it is similar to the HP-5082-3080 PIN diode that is used in the analyzer, and the spice model is available.
- I do not know what are the currents through those PIN-diodes, you would have to measure those and adjust the excitation in the simulation to get meaningful results. It is not only the current but the source resistance to CR1 PIN-didode that is also important, as it also influences the the over all gain (at least in the simulation).
- The tuning of PIN-ATTEN (C8) had significant impact on overall gain (depending on the source resistance of the CR1 current source). You might want to try to adjust that to see if it increases the gain for your amp (irrespective what the adjustment procedure tells you). But do mark the original position of C8 before you adjust it.
- The higher voltages on Q2 - Q5 emitters and bases are primarily set by R6 and R7. If anything, those are the resistors to start to check with. On the other hand those higher voltages that you measured only changed the gain of those stages just a little bit, less than 2dB in total, and so that is not the main source of your problems.

As mentioned, I would try to adjust the INPUT-MATCH, and PIN-ATTEN tuning if you could get more gain, after all this is a tuned amp!
The OUTPUT-MATCH does not have such a sharp peak.
If they were in that circuit before as you mentioned, it is conceivable that the tuning is off a bit.

I routinely use simulation for fixing broken equipment, I found it one of the best tools to explore a circuit.
If you are into LTspice simulation I could send along the simulation files and you could further explore the behavior of that amp.

Regards, Peter


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Sep Wed 01, 2021 6:40 pm 
Member

Joined: Mar Wed 16, 2011 10:44 pm
Posts: 2065
Location: Peekskill, NY
FWIW the hex-shaped 'terminals' in the lower photo are Murata Erie devices, either bypass capacitors or pi-section LC filters.
They're there to either keep RFI inside the box, or keep RFI outside the box.


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Sep Sat 11, 2021 3:12 pm 
New Member

Joined: Jan Mon 20, 2020 3:24 pm
Posts: 6
Hello readers,

To be able to test some components, the 50 MHz amplifier board had to be removed from the housing. Pins A2C8 to A2C13 are not connector pins but solder pins. The resistors R5 R6 R7 were good C4 was also good. The transistor Q2 was defective, it had a lot of leakage between base-emitter and base-collector. After replacing Q2, all DC voltages were according to the service documentation.

The gain was about 4 dB higher but still too low. The FET also becomes less warm. The real cause of the low signal was the wrong wiring between the PCB and the coax connector J2 and J4 . These were, probably during an earlier repair, reversed. The measuring circuit thus received the much lower signal from the AUXILARY output.

After this I adjusted the 50 MHz amplifier according to the 37 steps of the service documentation.

I want to thank everyone who contributed to this forum for their input.

Below you can see some pictures of the self-made extender cables that are needed for the repair of this spectrum analyzer.



Front and back sub-d connector. The coax connectors are fixed in a flexible paste.
Attachment:
sub-d-f-male.jpg
sub-d-f-male.jpg [ 583.5 KiB | Viewed 452 times ]



The total signal cable without the protective caps.

Attachment:
sig-kabel-tot.jpg
sig-kabel-tot.jpg [ 686.58 KiB | Viewed 452 times ]


Old connectors are used for the power extender cables.
Attachment:
voeding-f-male.jpg
voeding-f-male.jpg [ 519.86 KiB | Viewed 452 times ]

Attachment:
voed-male.jpg
voed-male.jpg [ 504.07 KiB | Viewed 452 times ]

Attachment:
voed-totaal.jpg
voed-totaal.jpg [ 1021.65 KiB | Viewed 452 times ]



Measurement setup with extender cables
Attachment:
totaal-kabels-toegepast.jpg
totaal-kabels-toegepast.jpg [ 937.03 KiB | Viewed 452 times ]


Measurement setup for adjustment 50 MHz amplifier with external 15 dB attenuator

Attachment:
meetopst-tot.jpg
meetopst-tot.jpg [ 929.21 KiB | Viewed 452 times ]


During the 37 steps when adjusting the 50 MHz amplifier, the resistors R11 and R13 have to be disconnected and connected a few times.
To perform this operation without soldering, two jumpers are provided.

Attachment:
jumper.jpg
jumper.jpg [ 934.87 KiB | Viewed 452 times ]


Top
 Profile  
 
 Post subject: Re: HP 8555A ADVICE NEEDED
PostPosted: Sep Sun 12, 2021 3:05 am 
Member
User avatar

Joined: Oct Sun 15, 2006 12:57 pm
Posts: 7964
Location: Liberty, Missouri
Loek, I'm Glad to see you have your problem resolved. I have a 141 with the TGs that I've used for years, and have repaired a few others as well. Crude by today's standards but still very useful.

_________________
Pics, Bold & underlined text, are usually links.

https://mikeharrison.smugmug.com/

Mike


Top
 Profile  
 
Post New Topic Post Reply  [ 10 posts ] 

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 26 guests



Search for:
Jump to:  




































Privacy Policy :: Powered by phpBB