8753ES YIG oscillator circuit WCA
One of the most challenging troubleshooting endeavors is when there is a misbehaving analog circuit where insufficient or vague limits are specified in the service documentation.? I have such a situation with my 8753ES in the phase lock circuit.
The pretune test waveform shown in the service guide differs from what I see by an offset.? A replacement phase lock board also has such an offset, and it came in a sealed envelope from the factory.
A complicating factor is that the instrument will sometimes "automatically" recalculate the pretune values when there is a failure detected during sweep, so by the time you can get to the test waveform you cannot be entirely sure if you are seeing the default pretune values, corrected pretune values, or partially corrected pretune values after a failed internal cal routine (test 48).? The best way to find out is to go into source tune mode where the instrument displays the DAC values and tune to 300 kHz.? If you see a DAC value of 1019 then at least at that frequency you are seeing the default value.
I decided I would do a Worst Case Analysis of the drive circuit when the DAC is set to 1019 (decimal).? If anyone is interested in the details I can discuss that but a summary is as follows:
1)? I assumed no selected values, that all components are of the tolerance in the HP parts list for the board as shown in the 8753D CLIP (same board part number used in the 8753ES).
2)? The 6.2 volt reference Zener has a tolerance of 2%.
3)? The voltage shift and buffer amps use resistors with 0.1% tolerance.
4)? The current sense resistor is 1%.
5)? The current to voltage converter is closed loop and has negligible error.
6)? The DAC has a +/- 1 LSB error, and at 12 bits, that's about 0.025% so I ignored it in this analysis.
7)? The YIG oscillator per Vladan is 23 +/- 1 mA/GHz.
8)? Component tolerances were arranged to form worst case combinations to make the extremes and still be at the edge of spec.
9)? I assumed the 3.8 GHz oscillator had no error.? The tolerance of this will add to these WCA numbers.
The end result is that the uncorrected output frequency of the source using the default DAC pretune value at a setting of 300 kHz will be -217 to +368 MHz.? Note the negative value.? The firmware will have to comprehend that increasing the DAC value will cause a decrease in output frequency and act accordingly.
The pretune 1 V/GHz output at the default DAC setting of 1019 can be from 3.474 to 3.631 volts with a nominal of 3.552 volts.? Thus my apparently shifted output is completely normal and in spec.? Those who see lower shifted outputs are most likely seeing corrected values where the DAC is no longer at 1019.? Both of my phase lock boards are well within these limits with the largest error contribution being the Zener.? My YTO, after doing a more careful series of measurements, comes out to be 22.251 mA/GHz, also within spec.? I took measurements of the second YTO and they look close but haven't done the curve fit to know the actual sensitivity.
So this leaves me with the same mystery as when I started:? why is my unit failing test 48 when all the tests in the troubleshooting guide seem to give good results?*
(* with some slight exceptions depending on how the tests are performed)
Peter
|
Re: HP5335A fan replacement
Dr. David Kirkby from Kirkby Microwave Ltd
LOn Tue, 19 Mar 2019 at 8:18 pm, Dan Nessett via Groups.Io <dnessett= [email protected]> wrote: The specs for the?Pabst fan that comes with the HP5335A are:?34 dBA, 61 m3/hour. The specs for the Evercool fan are: 28 dBA, 55 m3/hour. So, the Evercool is a bit underpowered, but considerably quieter than the?Papst. From what I have read the?Papst fan is a bit?over spec'd in?terms of airflow, so I thought the Evercool fan might work fine.
The flow rates you quote are probably in free air. Back pressure will reduce that considerably. You can pretty much ignore the free air figures. They should impressive, but have little connection with reality.
You will detect a 6 dB difference in noise, but the difference will not be considerable. It will not appear 4 times as loud. In fact, a 3 dB difference is just about percectable.?
Dave. -- Dr David Kirkby Ph.D C.Eng MIET
Kirkby Microwave Ltd
Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, CHELMSFORD, Essex, CM3 6DT, United Kingdom.
Registered in England and Wales as company number 08914892
Tel 01621-680100 / +44 1621-680100
|
Re: 8753ES YTO freq vs. Main Coil mA
Hi Peter,
Your plot shows? y = 0.0444x + 0.0656. So, there is a 65.6 MHz offset at 0 mA main coil current. This could in part be caused by a FM coil current which does not show up across the 40 Ohm resistor. Basically, I think you have confirmed that your YTO works and produces an output well below 3.8 GHz. The only exception to this statement would be a case where the YTO stops working in a narrow range of frequencies, but that almost always happens near the very low end of the oscillator's range (well below 3.8 GHz).
The question now is, how to allocate the 65.6 MHz offset. How much of it is due to some FM coil current and how much is due to some other problem. Keep in mind what I said about the FM coil current, it should not exceed 150 mA, which corresponds to approximately 35 MHz tuning range in either direction. I guess, I would somehow check to make sure the coil isn't being fried with too much current.
I have no idea how the firmware deals with pretune constants in CW mode and/or PLL off.
Vladan
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
Hi The battery backuped ram (M48Z18) in my HP 54645 died on me.. I desoldered it, put in a socket.? Then I fixed the ram by heating it a bit, and pry it open. Removed the old battery, and soldered on a big lithium cell, looks like the same type as seen 90's PCs. See picture... Should last a very long time.?
Br, Askild
On Sun, 24 Mar 2019, 17:47 Reginald Beardsley via Groups.Io, <pulaskite= [email protected]> wrote: The Tek 11801 has 4 Panasonic BR 2/3 A? 1200 mAh? 3 V lithium cells.? All of those tested at 3.07 to 3.1 V despite manufacturing dates of 12-87 and 09-88.? I think the fundamental issue with the NVRAM is the cells are too small.? The ST? battery caps for the SMD parts have a choice of 40 or 120 mAh capacities.
Once I get my replacements I'll try to reverse engineer the circuit in the NVRAM sockets.
|
Re: My $25 "In Poor Shape" 3468A has arrived
On Sat, Mar 23, 2019 at 10:49 AM, ghnatiuk@... wrote:
Alex,
PayPal is fine.? Send to ghantiuk@...?and send as a friend so PayPal does not extract money out.
Stranded wire will not work.
If you want two jumpers for 4-wire Ohm cal(SOURCE and SENSE), then add and additional $5.
State a complete address with city etc.? I will not replace items if the mail messes up.? There is a limit to my time in making these jumper shorts.? The ends have to be adjusted to the proper diameter.? #6 solid wire is close but not on the money. The price I am giving you is Waayyyyyyyyy below cost in material and labor.
GH
?Hi George, Just sent you $10 via Paypal, and follow up email with address as sending money to a friend does not allow to include one. Thanks! Alex
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
The Tek 11801 has 4 Panasonic BR 2/3 A 1200 mAh 3 V lithium cells. All of those tested at 3.07 to 3.1 V despite manufacturing dates of 12-87 and 09-88. I think the fundamental issue with the NVRAM is the cells are too small. The ST battery caps for the SMD parts have a choice of 40 or 120 mAh capacities.
Once I get my replacements I'll try to reverse engineer the circuit in the NVRAM sockets.
|
Re: I'm officially lost now with this 8753ES
Hello,
with apologies, I need to concur and agree.
There indeed is a header in the specification. You learn something every day.
Greets from Budapest, 3m under the ground
- - - - -
With best regards
Tam HANNA
Enjoy electronics? Join 13500 followers by visiting the Crazy Electronics Lab at
toggle quoted message
Show quoted text
On 23.03.19 22:50, Dave McGuire wrote: On 3/23/19 1:42 PM, Tam Hanna wrote:
Please let me protest the designation of BMP as a Windows format. I was referring to its origins, and the platform on which it sees by
far the most application. It was developed by Microsoft, for use within
Windows, in the mid-1980s. It has long been supported on most other
platforms that have graphical capabilities, but that changes neither its
origins nor its primary platform.
A BMP, short for BitMaP, simply is the content of a framebuffer farted out to a
disk. Not quite. A BMP file contains a header with metadata (actually two
such headers), then a colormap, then pixel data.
-Dave
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
I like the discreet component approach here and it looks like it would work well for Tektronix 2465B scopes as well that have a severe height constraint.? With respect to alkalines H.. No. Bad idea and I have done it to myself. The manufacture and quality of any alkaline these days seems seriously questionable.? Regards Paul. WB8TSL
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
That would be an interesting solution but they're actually a different lithium chemistry.? An iron disulfide cathode gives the 1.5 volt cell voltage:
Peter
toggle quoted message
Show quoted text
On 3/24/2019 11:47 AM, Chuck Harris wrote: As far as I know the 1.5V lithium AA cells are a hybrid between
a LDO power regulator, set to 1.5V and a 3V lithium cell of some
sort. The few I have found put out somewhere around 1.8V unloaded,
and drop back to 1.5 under load. I have been meaning to open one up.
I am not sure how they would hold up under a no current draw situation
like backing up a CMOS RAM.
-Chuck Harris
Peter Gottlieb wrote:
And they seem to be getting worse all the time.
What about those new lithium 1.5 volt AA cells? Very long shelf life and I have yet to see one leak. Seems like an ideal choice for memory backup.
Peter
On Mar 24, 2019, at 10:45 AM, Daun Yeagley <daun@...> wrote:
I couldn't agree more! I've had *terrible* experiences with alkaline cells leaking.
Daun
On 3/24/2019 10:25 AM, Chuck Harris wrote:
I suggest in the strongest terms, do NOT use alkaline
cells in any memory backup scheme. They will leak out
their electrolyte long before they run out the necessary
power to run the backup scheme.
I have done this experiment too many times, and lost too
much equipment to the damage the alkaline cells have done.
If you simply must use alkalines, put them on a long leash
and mount them outside of the instrument.
-Chuck Harris
Jack2015 wrote:
Usually there are three parts in a NVRAM,
1.SRAM
2.DS1210(or DS121x?)
3. 3V battery
This means we can make NVRAM by ourself,
HM6116+DS1210+two 1.5V AAA batteries=DS1220Y(2Kx8 NVRAM)
one 256K SRAM+DS1210+two 1.5V AA batteries(or one 3V battery)=DS1230Y=M48Z35Y 32kx8 NVRAM,
couple years ago I made following NVRAM for tek 11A32,last week I replaced the HM6116 with a 256K SRAM
and used it in CSA803(for a M48Z35Y=DS1230Y).it works well.by using larger capacity battery the life of data
storage will be much much longer!
Attachments:
NVRAM.jpg: /g/HP-Agilent-Keysight-equipment/attachment/95351/
--
Daun E. Yeagley II, N8ASB
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
As far as I know the 1.5V lithium AA cells are a hybrid between
a LDO power regulator, set to 1.5V and a 3V lithium cell of some
sort. The few I have found put out somewhere around 1.8V unloaded,
and drop back to 1.5 under load. I have been meaning to open one up.
I am not sure how they would hold up under a no current draw situation
like backing up a CMOS RAM.
-Chuck Harris
Peter Gottlieb wrote:
toggle quoted message
Show quoted text
And they seem to be getting worse all the time.
What about those new lithium 1.5 volt AA cells? Very long shelf life and I have yet to see one leak. Seems like an ideal choice for memory backup.
Peter
On Mar 24, 2019, at 10:45 AM, Daun Yeagley <daun@...> wrote:
I couldn't agree more! I've had *terrible* experiences with alkaline cells leaking.
Daun
On 3/24/2019 10:25 AM, Chuck Harris wrote:
I suggest in the strongest terms, do NOT use alkaline
cells in any memory backup scheme. They will leak out
their electrolyte long before they run out the necessary
power to run the backup scheme.
I have done this experiment too many times, and lost too
much equipment to the damage the alkaline cells have done.
If you simply must use alkalines, put them on a long leash
and mount them outside of the instrument.
-Chuck Harris
Jack2015 wrote:
Usually there are three parts in a NVRAM,
1.SRAM
2.DS1210(or DS121x?)
3. 3V battery
This means we can make NVRAM by ourself,
HM6116+DS1210+two 1.5V AAA batteries=DS1220Y(2Kx8 NVRAM)
one 256K SRAM+DS1210+two 1.5V AA batteries(or one 3V battery)=DS1230Y=M48Z35Y 32kx8 NVRAM,
couple years ago I made following NVRAM for tek 11A32,last week I replaced the HM6116 with a 256K SRAM
and used it in CSA803(for a M48Z35Y=DS1230Y).it works well.by using larger capacity battery the life of data
storage will be much much longer!
Attachments:
NVRAM.jpg: /g/HP-Agilent-Keysight-equipment/attachment/95351/
--
Daun E. Yeagley II, N8ASB
|
Re: 8510C 85102A IF/DETECTOR L.O. Phaselock Out Upgrade / Option ?
My 8510C, with no listed options, has the connector too.
It is isolated between a pair of fiberglass washers, and
is in the cluster of connectors on the back of the IF/Detector
unit.
I guess what I am having trouble parsing from the OP's statement
below, is:
Is the hole the cable is coming out of labeled L.O. Phaselock Out,
or is the cable marked L.O. Phaselock Out?
Is the added hole just drilled somewhere on the back panel, or is
it one that was already there from the factory?
The way I am reading it is: There is a miscellaneous hole drilled
somewhere on the back panel of the IF/Det unit that has been labeled
with L.O. Phaselock Out, and it has a cable hanging out of it isn't
identified as to its purpose.
-Chuck Harris
Martin Rickes wrote:
toggle quoted message
Show quoted text
Hi Kevin,
at my 8510C there's also such an output. Mine looks quite original to me,
and my unit was Agilent-cal-sealed, when I got it.
Maybe somebody did brake the original BNC (it is fully isolated from the
case -> see picture) on yours and they bodged something together?
As you see, I did not connect it to anything (83620-Source).
73
Martin, DL8RI
Am So., 24. M?rz 2019 um 15:18 Uhr schrieb Kevin Hobbs <ve3kh@...>:
Hi 8510 Experts
I am piecing together an 8510C System ... the 85102A IF/DETECTOR has a BNC
cable hanging out of an added rear panel hole with a label that says "L.O.
Phaselock Out" ... Does anyone recall an upgrade or Option like this and
how it connects to the rest of the system etc?
Any wisdom appreciated !
73 Kevin
Attachments:
8510C_Phase_lock_bnc.jpg: /g/HP-Agilent-Keysight-equipment/attachment/95356/0
|
Re: HP5335A fan replacement
I've read that HP sized its fans to cool the instrument even when operating at maximum outside air temperature. If that is true, one might get away with less airflow if ambient temps stay closer to "normal room temperature." My two 5245-family counters are definitely in need of quieter fans. I've thought of replacing the stock 4-inch fan with a 3-inch fan mounted on a heavy rubber sheet. The sheet would provide mounting, air control, and vibration isolation. Obviously that would not work on smaller, newer instruments that use a 3-inch fan to begin with.? ---Jeremy
On Sun, Mar 24, 2019 at 2:34 AM Andy ZL3AG via Groups.Io <zl3ag= [email protected]> wrote:
Would this be a situation where the best option would be to use some CPU cooler fans to direct local airflow to the hot components, allowing quieter fans to do the job in the back frame?
On 24/03/19 6:12 PM, Dan Nessett via Groups.Io wrote:
> I have looked high and low for a 80x80 fan that has equal airflow and lower noise specs that the fan already installed in my HP5335A without success. Given the comments by Joe / KN5U that many of the units he has repaired had replacement fans with lower airflow specs and failed Schmitt amplifiers (parts that are no longer available), I have decided to stick with the (noisy) installed fan.
|
Re: HP5335A fan replacement
Hello to the group. The old noisy bad bearing fan issue. I agree very hard to find replacements these days. Hard enough that I have started to open the fans up and repair them. I am seriously a beginner at this. But it seems the fans are technically easy to disassemble. (Thats theory) Almost always there are little c clips you have to be careful with. They are small but key to getting in. I use various tools to release and then put them back in. Absolutely there should be official tools but when looking on epay as an example I can never figure out exactly what ones. When in stores that carry such things the points never seem small enough.
?With respect to sleeve bearings. After getting the fan assembly apart I clean it up. Transmission oil seems to do a nice job of getting rig of the old grease and dirt. Then I use gun grease to re-lube the bearings.
There may be better and lighter grease. Just hard to say whats correct.? I have not ventured into replacing real bearings yet. But that will happen and there really seems to be a lot of bearings available online. My understanding is these are press fit and can be knocked out. One day I will find out.
I suppose I should get ahead of that learning curve and buy some old fans at the next ham flea market and just experiment.
Regards Paul WB8TSL
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
And they seem to be getting worse all the time.
What about those new lithium 1.5 volt AA cells? Very long shelf life and I have yet to see one leak. Seems like an ideal choice for memory backup.
Peter
toggle quoted message
Show quoted text
On Mar 24, 2019, at 10:45 AM, Daun Yeagley <daun@...> wrote:
I couldn't agree more! I've had *terrible* experiences with alkaline cells leaking.
Daun
On 3/24/2019 10:25 AM, Chuck Harris wrote:
I suggest in the strongest terms, do NOT use alkaline
cells in any memory backup scheme. They will leak out
their electrolyte long before they run out the necessary
power to run the backup scheme.
I have done this experiment too many times, and lost too
much equipment to the damage the alkaline cells have done.
If you simply must use alkalines, put them on a long leash
and mount them outside of the instrument.
-Chuck Harris
Jack2015 wrote:
Usually there are three parts in a NVRAM,
1.SRAM
2.DS1210(or DS121x?)
3. 3V battery
This means we can make NVRAM by ourself,
HM6116+DS1210+two 1.5V AAA batteries=DS1220Y(2Kx8 NVRAM)
one 256K SRAM+DS1210+two 1.5V AA batteries(or one 3V battery)=DS1230Y=M48Z35Y 32kx8 NVRAM,
couple years ago I made following NVRAM for tek 11A32,last week I replaced the HM6116 with a 256K SRAM
and used it in CSA803(for a M48Z35Y=DS1230Y).it works well.by using larger capacity battery the life of data
storage will be much much longer!
Attachments:
NVRAM.jpg: /g/HP-Agilent-Keysight-equipment/attachment/95351/
--
Daun E. Yeagley II, N8ASB
|
Re: 8510C 85102A IF/DETECTOR L.O. Phaselock Out Upgrade / Option ?
Hi Kevin,
at my 8510C there's also such an output. Mine looks quite original to me, and my unit was Agilent-cal-sealed, when I got it.
Maybe somebody did brake the original BNC (it is fully isolated from the case -> see picture) on yours and they bodged something together?
As you see, I did not connect it to anything (83620-Source).
73 Martin, DL8RI
Am So., 24. M?rz 2019 um 15:18?Uhr schrieb Kevin Hobbs < ve3kh@...>:
Hi 8510 Experts
I am piecing together an 8510C System ... the 85102A IF/DETECTOR has a BNC cable hanging out of an added rear panel hole with a label that says "L.O. Phaselock Out"? ... Does anyone recall an upgrade or Option like this and how it connects to the rest of the system etc?
Any wisdom appreciated !
73 Kevin
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
I couldn't agree more!?? I've had *terrible* experiences with alkaline cells leaking.
Daun
toggle quoted message
Show quoted text
On 3/24/2019 10:25 AM, Chuck Harris wrote: I suggest in the strongest terms, do NOT use alkaline
cells in any memory backup scheme. They will leak out
their electrolyte long before they run out the necessary
power to run the backup scheme.
I have done this experiment too many times, and lost too
much equipment to the damage the alkaline cells have done.
If you simply must use alkalines, put them on a long leash
and mount them outside of the instrument.
-Chuck Harris
Jack2015 wrote:
Usually there are three parts in a NVRAM,
1.SRAM
2.DS1210(or DS121x?)
3. 3V battery
This means we can make NVRAM by ourself,
HM6116+DS1210+two 1.5V AAA batteries=DS1220Y(2Kx8 NVRAM)
one 256K SRAM+DS1210+two 1.5V AA batteries(or one 3V battery)=DS1230Y=M48Z35Y 32kx8 NVRAM,
couple years ago I made following NVRAM for tek 11A32,last week I replaced the HM6116 with a 256K SRAM
and used it in CSA803(for a M48Z35Y=DS1230Y).it works well.by using larger capacity battery the life of data
storage will be much much longer!
Attachments:
NVRAM.jpg: /g/HP-Agilent-Keysight-equipment/attachment/95351/
--
Daun E. Yeagley II, N8ASB
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
I suggest in the strongest terms, do NOT use alkaline
cells in any memory backup scheme. They will leak out
their electrolyte long before they run out the necessary
power to run the backup scheme.
I have done this experiment too many times, and lost too
much equipment to the damage the alkaline cells have done.
If you simply must use alkalines, put them on a long leash
and mount them outside of the instrument.
-Chuck Harris
Jack2015 wrote:
toggle quoted message
Show quoted text
Usually there are three parts in a NVRAM,
1.SRAM
2.DS1210(or DS121x?)
3. 3V battery
This means we can make NVRAM by ourself,
HM6116+DS1210+two 1.5V AAA batteries=DS1220Y(2Kx8 NVRAM)
one 256K SRAM+DS1210+two 1.5V AA batteries(or one 3V battery)=DS1230Y=M48Z35Y 32kx8 NVRAM,
couple years ago I made following NVRAM for tek 11A32,last week I replaced the HM6116 with a 256K SRAM
and used it in CSA803(for a M48Z35Y=DS1230Y).it works well.by using larger capacity battery the life of data
storage will be much much longer!
Attachments:
NVRAM.jpg: /g/HP-Agilent-Keysight-equipment/attachment/95351/
|
8510C 85102A IF/DETECTOR L.O. Phaselock Out Upgrade / Option ?
Hi 8510 Experts
I am piecing together an 8510C System ... the 85102A IF/DETECTOR has a BNC cable hanging out of an added rear panel hole with a label that says "L.O. Phaselock Out"? ... Does anyone recall an upgrade or Option like this and how it connects to the rest of the system etc?
Any wisdom appreciated !
73 Kevin
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
Usually there are three parts in a NVRAM, 1.SRAM 2.DS1210(or DS121x?) 3. 3V battery This means we can make NVRAM by ourself, HM6116+DS1210+two 1.5V AAA batteries=DS1220Y(2Kx8 NVRAM) one 256K SRAM+DS1210+two 1.5V AA batteries(or one 3V battery)=DS1230Y=M48Z35Y 32kx8 NVRAM, couple years ago I made following NVRAM for tek 11A32,last week I replaced the HM6116 with a 256K SRAM? and used it in CSA803(for a M48Z35Y=DS1230Y).it works well.by using larger capacity battery the life of data storage ?will be much much longer! 
|
Re: NVRAM 28-Pin Encapsulated Package DS1230Y and 6811B
I just ran into a pair of standard NEC D43256 SRAMs in a socket which contains a small board and two small cells in a 1991 vintage Tek 11801 I am restoring to operation from an E5622, dead NVRAM battery, error.
|
Jeremy Nichols
Martin Rickes
Kevin Hobbs