开云体育

To avoid trampling over another group bandwidth i suggest you as this on the Marconi group.io pages

I have a 3852A with firmware 3.0 and wanted to use a card that requires firmware 4.2 . I found the instructions and firmware 4.22. I identified my board as 66503 revA? per PCB . Per instructions I only needed to hook pins 1 of U57 and U58 to pin 4 of RP4. I did that but I get a error message that the ROM update is not complete.?
Anyone with revA of 66503 board had this happen?
The suggested remedy on the display is not possible on Rev A boards as there is no jumper wire .?

PAI had a minimum order of 100, but when I went up to 500 they started to seem really worth it.? The worst thing about doing a bulk buy was having to count out all those little bags of 50 :^)? I've had no complaints since.
Dan

I asked Precision for a quote, I guess we will see Monday.

Wally

Many thanks David, somehow i missed that in my search.

Regards Pete

开云体育

开云体育

开云体育

I have the latest version of my PCB's ready to build. I added a trimpot for the bias on the prescaler chip to do away with the need for selecting a resistor, which is accessible through a convenient spare hole in the side of the chassis.
This way you can easily dial in the bias to optimise the tradeoff between stopping random counts with no input signal and sensitivity of the prescaler.

Here is the sensitivity of my previous version (pdf), once I have my latest version constructed, I'll make a better measurement of its performance.
It exceeds the HP spec by a bit up the top of the frequency range, so I'll have to figure out how to fix that.


And a 3D render of the design:

Since the problem appears to be a residual reading at the lowest ranges and the grid of V2 helps, you might want to rule out spurious noise getting to V2. Multi MHz oscillation in the V1 stage comes to mind.
I'd check (with a 50 MHz or better 'scope) at the grid of V2. The "phase of the moon" observation supports this hypothesis. Perhaps one or more of the capacitors in the V1 circuit is bad.

Have a look for the 400D & 400L on here too, they are almost identical to the 400H.

Looking at the manual, it lists a number of paper capacitors, these will almost certainly be leaky, some are grid coupling capacitors.
Also from experience I've had a number of the 1500uF low voltage capacitors fail open circuit, these were single cans rather than the dual can used in this, they were in a pair of 412A and another in a 5245L.

David

Greetings to the Group:

?? I may have botched the search function, as I can't find any actual discussions of repairs to this instrument.?? If so, I apologize for taking up the bandwidth.

?? However, I do seem to need a bit of help with this instrument.?? The unit will not zero on the 1 volt / 0 dB scale; it reads anywhere from about 1/3 scale to 2/3 scale, depending on phases of the moon or whatever.?? The reading decreases proportionally on the higher voltage scales.?? On the .3 volt / -10 dB scale, the zero is pretty close, and remains so on the more sensitive scales.

?? On the .3 volt / -10 dB scale, the meter cannot be calibrated; with R29 at the maximum position and .3 volts applied, the reading is .18 volts.?? R29 seems to have little effect on the meter level.?? It can be used to bring the meter to read .3 volts on the 1 volt scale as per the manual. but the meter reads only slightly less with no signal applied and the input terminals shorted.

?? The manual suggests that the trouble is in the power supply, and indeed, I see a 60 Hz signal in the last two stages of gain (V4 and V5).?? Unfortunately, the power supply looks very clean with my oscilloscope.?? I note that there is only 3 mV of ripple permitted, so I had to use a series .1 uF capacitor to protect my scope from the B+ when using a X1 probe.?? At 5 mV per division, I see only about 1 mV (barely discernible).?? I have two of these instruments and have them opened up side by side on the bench.?? One works well and the other is failed as described.?? The two instruments have very similar power supply ripple, with the good instrument possibly being a bit worse than the failed one.

?? Filament supply DC ripple is more pronounced, but is no worse than (and may be slightly better if I recall correctly) than the working unit.

?? Grounding the grid of V2 gets rid of the noise and allows the instrument to zero.?? V1 and V2 have both been replaced with no improvement.?? All of the tubes have been pulled from the instrument and checked with a TV-7 B/U tube tester.?? All check good for shorts, mutual conductance and gas.

?? I am looking for any suggestions at this point.... I seem to be completely stumped.

? Thanks for reading and for nany suggestions you may have.

73,
--
Jim Thorusen
KB6GM

The HP 8900B manual with circuits diagrams is available from BAMA;

David

The CPU board slides out from the back. The ram slot is available through a cutout on the CPU board. The CPU is a 68000 and HP made only 256k and 1M cards for these while other vendors made 2M and 4M versions. I think other card where not available due to ram cost.? Not sure that the system can handle more then 4M but the CPU is able to address up to 16M of address space.?

Razvan:

The 3GHz board is a divide by 128, with a max output of 23.4375MHz at 3GHz.
That has to get to the main board through the ribbon cable, not the best way to connect.
At 12GHz the output would be almost 100MHz so I think that the higher frequency option boards likely have an additional divider to keep the output at a reasonable frequency.

There are two pins (CH3CODE0 and CH3CODE1) on the connector that are used to tell the main board that one of the options is installed.

I would think that a combination of those pins would select one of the three options or no option.

I will try to see what effect they have later this weekend.

ed

Yves:

Thanks for the update.

Do the boards count randomly when there is no input?

Were these the OEM options or clones?

ed

开云体育

In metric-land I got a bunch of 1mm ID 2mm OD watch stem seals, eBay or Amazon I seem to recall. That was about 10 years back on an 8566A - still working fine - done several other attenuators since.
However while they seem OK at the moment, nowadays having had cheap plumbing 'rubber' seals go gooey after about 5 years, I have little faith in unbranded stuff and I'll bet it's just a matter of time until they start to return to their original liquid hydrocarbon form.
Probably worth finding some of that size from a known (and trusted) manufacturer?

hello friends,

I got a 8593A with a unknown problem,
to me it looks to be a failes display board. Any tips to short the search ?
thank you.

see the short video I have made from that :-)


greetings from germany
Martin

开云体育