|
My pal Robert PE1MJA took the time to reverse engineer the MBF so all credits to him! And no guarantees there are no mistakes, goes without saying I suppose..
The trick is playing with the caps to peak the filter to 10 instead of 5 MHz.
hth, Wilko
|
On 9/22/24 17:50, Wilko Bulte wrote: My pal Robert PE1MJA took the time to reverse engineer the MBF so all credits to him! And no guarantees there are no mistakes, goes without saying I suppose..
The trick is playing with the caps to peak the filter to 10 instead of 5 MHz. Ahh, this is very handy, thanks Wilko. Please extend my thanks to Robert as well. -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
Am 22.09.2024 um 23:50 schrieb Wilko
Bulte:
My pal Robert PE1MJA took the time to reverse engineer the MBF so all credits to him! And no guarantees there are no mistakes, goes without saying I suppose..
I am a little bit surprised the outputs are not isolated.
Bernd
|
No, they are simple BNC that are bolted to a common faceplate. No transformers on the output, no isolated BNC etc.
Wilko
|
I recall there was more to it than just the capacitors.
I reversed the design, quite a long time ago, and this is what I came
up with... I cannot remember if I got any clues from the internet, but
I don't think I did, given my copious notes, and my proclivity for
tracing out schematics for gizmos I want to understand.
1) Remove all jumpers and 1K resistor from around HC390 chip.
2) remove HC390 from board.
3) fit 1K resistor in holes E & F using outer holes (away from HC390).
4) remove resonator caps from M.B. near toroid.
5) Pick new caps to resonate @ 10MHz. A starting point would be 1/4 the
capacitance used for 5MHz. I used 191pf.
6) Apply 20V to TP1 and TP2, 0.5V RMS to input connector (top
connector). Connect scope with 50 ohm load to output connector,
adjust for 0.5VRMS.
7) Amplifier stages should be set for 0.5VRMS, 50 ohm.
8) Fault pot should be set to illuminate LED at 0.2VRMS.
I have a hand drawn schematic of some of the MBFD, but I don't think
it is all that useful, unless you are trying to figure out what was
done.
I don't recall what I meant by "adjust" on step 6)... I think I meant
to adjust the input for an output on the scope of 0.5VRMS. I am guessing
it will become obvious if you have one of these modules opened up and on
your bench.
My change sets up the MBF to properly put out a big clean signal at
10MHz, and to have a functioning "error" indicator. I had no other info,
so I just guessed that 0.2VRMS was an appropriate threshold.
-Chuck Harris
toggle quoted message
Show quoted text
|
Most definitely we did not touch the 390's. And the error indicator LED also works. I should have notes about what it is exactly we did. But where...
Did find these two pictures, caps were taken from the MBF for sure ?
|
The HC390 I removed was being used as a divide by 2 to bring 10MHz down
to 2MHz. Because the input was a square wave, one shouldn't expect
there to be much if any 3nd harmonic, so if you didn't remove the divide
by 2, or do a cut and bypass job, there wouldn't be much 10MHz to feed
the filter and amplifier.
It is within the realm of possibilities that there are more than one
version of the MBF module out there.
-Chuck Harris
toggle quoted message
Show quoted text
|
2MHz -> 5MHz. Getting old sucks! -Chuck Harris On Wed, 25 Sep 2024 10:40:07 -0400 "Chuck Harris" <cfharris@...> wrote: The HC390 I removed was being used as a divide by 2 to bring 10MHz
down to 2MHz. Because the input was a square wave, one shouldn't
expect there to be much if any 3nd harmonic, so if you didn't remove
the divide by 2, or do a cut and bypass job, there wouldn't be much
10MHz to feed the filter and amplifier.
It is within the realm of possibilities that there are more than one
version of the MBF module out there.
-Chuck Harris
On Wed, 25 Sep 2024 15:36:24 +0200 "Wilko Bulte" <wkb@...>
wrote:
Most definitely we did not touch the 390's. And the error indicator
LED also works. I should have notes about what it is exactly we did.
But where...
Did find these two pictures, caps were taken from the MBF for sure
?
Attachments:
image0.jpeg:
/g/HP-Agilent-Keysight-equipment/attachment/147485/0
image1.jpeg:
/g/HP-Agilent-Keysight-equipment/attachment/147485/1
|
Not sure what module you have but my MBF are just that: distrib amps, not dividers.
They have no HC390, the only ICs present are two LM139. The FRK-HLN Rb that they came with is a 5 MHz unit. So the MBF were originally setup for 5 MHz. I added a x2 multiplier, and hence had to adapt the MBF to run at 10 MHz.
hth, Wilko
|
I think that the alternative seems to suck more.....
Keep on keeping on......
Harvey
toggle quoted message
Show quoted text
On 9/25/2024 4:48 PM, Chuck Harris wrote: 2MHz -> 5MHz. Getting old sucks!
-Chuck Harris
On Wed, 25 Sep 2024 10:40:07 -0400 "Chuck Harris" <cfharris@...>
wrote:
The HC390 I removed was being used as a divide by 2 to bring 10MHz
down to 2MHz. Because the input was a square wave, one shouldn't
expect there to be much if any 3nd harmonic, so if you didn't remove
the divide by 2, or do a cut and bypass job, there wouldn't be much
10MHz to feed the filter and amplifier.
It is within the realm of possibilities that there are more than one
version of the MBF module out there.
-Chuck Harris
On Wed, 25 Sep 2024 15:36:24 +0200 "Wilko Bulte" <wkb@...>
wrote:
Most definitely we did not touch the 390's. And the error indicator
LED also works. I should have notes about what it is exactly we did.
But where...
Did find these two pictures, caps were taken from the MBF for sure
?
Attachments:
image0.jpeg:
/g/HP-Agilent-Keysight-equipment/attachment/147485/0
image1.jpeg:
/g/HP-Agilent-Keysight-equipment/attachment/147485/1
|
My unit is an extended, double high rack version of Ball/
Efratom's MFS-267... Which is a Modular Frequency Standard,
system comprised of a MGPS, MRK rubidium standard, and a bunch
of MBF's all in a 2 unit high rack chassis.
The MGPS and MRK standard on my system has a 1pps and a 10MHz
output. The MBF system takes that output and divides it down
to 5MHz in 8, 4 output MBF distribution amplifiers. There are
also an equal number of 1pps outputs in the MBP distribution
system.
My MBF distribution amplifiers each take a 10MHz input into
the card edge connector end, and output 4, 5MHz signals,
through the BNC connectors on the panel end.
Each MBF has a red led to indicate a fault, on its panel, and
there is one mostly blank MBF panel to provide a fault LED on
the MFS front panel.
The divider in each MBF on my units can be set for 1, 2, 5, or 10MHz,
by jumpers. I removed the HC390 because I didn't think it wise to
use Ball/Efratom's apparent method of jumpering the CLK input to
one of the 'HC390's output FF's.
It would surprise me to find that the circuit board in your MBF unit
wasn't capable of full operation... But, I get surprised a lot.
-Chuck Harris
toggle quoted message
Show quoted text
On Thu, 26 Sep 2024 00:01:53 +0200 "Wilko Bulte" <wkb@...> wrote: Not sure what module you have but my MBF are just that: distrib amps,
not dividers.
They have no HC390, the only ICs present are two LM139. The FRK-HLN
Rb that they came with is a 5 MHz unit. So the MBF were originally
setup for 5 MHz. I added a x2 multiplier, and hence had to adapt the
MBF to run at 10 MHz.
hth, Wilko
|
On Wednesday 25 September 2024 04:48:49 pm Chuck Harris wrote: 2MHz -> 5MHz. Getting old sucks!
-Chuck Harris Yeah, but it sure beats the alternative... -- Member of the toughest, meanest, deadliest, most unrelenting -- and ablest -- form of life in this section of space, ?a critter that can be killed but can't be tamed. ?--Robert A. Heinlein, "The Puppet Masters" - Information is more dangerous than cannon to a society ruled by lies. --James M Dakin
|
OK, my MBF are a different kettle of fish: no dividing going on, just amplification.
They were originally fed 5 MHz from the FRK-HLN Rb. Outputs are 5 individually output level adjustable sine waves. 4 are BNC on the front panel, 1 is on the DIN41612 connector at the rear of the MBF for use in the chassis. In original state the MBF were nicely peaked at 5 MHz. No way to get a 10 MHz through. Hence the playing with the filter. I also have a variant that has no front BNC, rather teflon coax runs to the DIN41612. This MBF is for chassis-internal distribution.
Picture of my MBF. A close look shows the 5 identical amps. Adjustment pots are on the rear of the PCB.
Aside: this Efratom was a deal you do not often get. It had dual FRK-HLN 5 MHz Rb, an MGPS, dual PSU, dual lead-acid battery backup modules, a phase meter to adjust the free running Rb against the MGPS disciplined one, a switchable output module to select one or the other Rb for driving the 6 MBF, a divider/synth module that produces 2.048 MHz and a monitoring module. All housed in a stacked 2x 3U and double-deep 19" module. Total cost €75
Came from a military mobile sat groundstation.
Wilko
|
Hi Wilke,
Ok, you spurred me into action. I took a close look at my
Ball/Efratom MFS, and found that I have two different types
of modules labeled MBF: one is on the front panel, with only
a Fault LED, and the other has multiple examples on the back
panel labeled MBF, but with a Fault LED, and 4 BNC connectors.
The front panel MBF is labeled on the side: MBFD, for Divider,
I would suppose.
The rear panel MBFs have no filter, and a couple of the transistors
jumpered out.
The front panel MBFD, has a little daughter board, plugged into
the dual row header marked J1, a filter, and all of its amplifier
outputs brought back to the card edge connector.
So, as I remembered, my MBF's are broadband, working equally well
at 5, or 10MHz.
I have enclosed pictures that should explain everything.
-Chuck Harris
toggle quoted message
Show quoted text
On Fri, 27 Sep 2024 08:06:55 +0200 "Wilko Bulte" <wkb@...> wrote: OK, my MBF are a different kettle of fish: no dividing going on, just
amplification.
They were originally fed 5 MHz from the FRK-HLN Rb. Outputs are 5
individually output level adjustable sine waves. 4 are BNC on the
front panel, 1 is on the DIN41612 connector at the rear of the MBF
for use in the chassis. In original state the MBF were nicely peaked
at 5 MHz. No way to get a 10 MHz through. Hence the playing with the
filter. I also have a variant that has no front BNC, rather teflon
coax runs to the DIN41612. This MBF is for chassis-internal
distribution.
Picture of my MBF. A close look shows the 5 identical amps.
Adjustment pots are on the rear of the PCB.
Aside: this Efratom was a deal you do not often get. It had dual
FRK-HLN 5 MHz Rb, an MGPS, dual PSU, dual lead-acid battery backup
modules, a phase meter to adjust the free running Rb against the MGPS
disciplined one, a switchable output module to select one or the
other Rb for driving the 6 MBF, a divider/synth module that produces
2.048 MHz and a monitoring module. All housed in a stacked 2x 3U and
double-deep 19" module. Total cost €75
Came from a military mobile sat groundstation.
Wilko
Attachments:
camphoto_1932422408.jpeg:
/g/HP-Agilent-Keysight-equipment/attachment/147531/0
|
I just looked at the MBF module board I started experimenting with a few years ago. I got a small rack part of some Efratom system with a PS, four or five MBFs, and a couple other modules, all made for 5 MHz. I wanted to convert them to 10 MHz, but never could find schematics or other info. One thing I found online was reference to a version that had a board layout that could fit a Minicircuits brand BPF module, or the equivalent made with discrete LC circuitry. I found no details on this.
?
I checked my MBF, and found 10 MHz simply would not go through them, so they clearly have very effective BP/LP filtering. I did some reverse engineering too, kind of like what Wilko put up recently. It took a while to figure out how the BPF works, and what actually tunes it. Then I started experimenting and gathering a collection of mica caps in the right range for tweaking. After that, the 10 MHz came through, but still awfully low in level. It turned out that on my units, there's a low-pass LC pi filter on each output, right at the end, so more tweaking, but in four places instead of one for the main BPF in the common preamplifier stage.
?
My MBF modules have the usual quad outputs to BNCs (plus one internal channel), with a volume control for each. The board looks essentially identical to those pictured from Wilko and Chuck, except that the spots C9/L3/C3 through C12/L6/C6 are populated with parts for the LPFs. Note that in the ones from Wilko and Chuck, the empty L positions are shorted by thin runs, which are cut if an L is used. I can't remember for sure, but apparently I was looking at changing both the Ls and Cs in proportion, to get similar characteristics at 10 MHz, or just dropping the C value. My notes indicate the original nominal values were 1.8 nF/1.8 uH/1.8 nF, and I changed the Cs to 910 pF. On a different channel, I had 1 uH instead, and was changing the C values. That's about where I left off on this project, so I still need to perfect the changes, scrounge up enough parts, then replicate it in all the modules.
?
BTW this module is part numbered 104702-001.
?
Ed
|
Dave McGuire
