Author Topic: Repairing and Upgrading an old PM3320A - Including adding FFT! (Read 50130 times)

SaabFAN · « **Reply #125 on:** November 13, 2017, 01:14:10 am »

I think I have reached the end of what I can do with the Scope at this point.
I have reduced the wait-time for DataTransferAcknowledge to about 130 - 140ns (delay-timer + propagation-delays in the logic-gates) and there hasn't been any real improvement. I could try to reduce the delay even further, but I think the improvement will be marginal at best and reducing the delay to anything below 80ns is most likely impossible because of the enable-times and propagation-delays of the address and data-buffers.

Getting even more speed than now requires a different CPU that can be clocked at higher speeds than 10.50 MHz. There are 25MHz rated ones available on ebay, but they are in a PLCC-Package, which would require a custom adaptor and probably the removal of the PGA-68 Socket due to height-restrictions inside the card cage. Something I'm not comfortable with.

Btw. has anyone of you guys had more luck finding a datasheet for the bus arbiter? If there is, it would be possible to cram the entire Processor-Board into an FPGA or CPLD and put everything on an "accelerator-board". That would be pretty awesome.

wazoo · « **Reply #126 on:** November 30, 2017, 01:50:04 pm »

hi all

SaabFAN mention me this thread because we have a PM3323

past year I recue it from road to the trash from my workplace , because of "unused / stored for years / out of calibration"

today we will let it give it a chance

thanks to the great job of SaabFAN

for the first power up we saw a completely noise screen, without test signal. bootup test relays sound and text screen looks fine.
after replacing 90% of axial capacitor with radial version , and all SMD capacitor in P2CCD board, we can see a noise but clear test signal.

oohhh good news!!! can we repair it ?? it looks like we are on the good way.

thankU SaabFAN for your hard job. we will follow your steps.

first power up....

after cap replacement...

capacitor removed with leakage near

new capacitor and cleaned section

some axial cap replaced with radial version

wazoo · « **Reply #127 on:** November 30, 2017, 02:40:21 pm »

wwwooowwww SaabFAN it is really fantastic and very hard your work.

finally I have finished reading all thread

our scope is in my brothers home.
I will try to make 4K photos

to show you how this unit is inside. and the ROM versions.

Some years ago, when I was a bit younger

, I liked to "overclock" my first 486 computer that starts from 25Mhz and changed crystal to 33Mhz.

This processor is older tecnology, I know, but it's fun to hear that you are trying to squeeze the processor some ns

I am not an engineer but can see that this oscope is really a "heavy" tool

we will follow yours and macboy colleague progress

this is our PM3323 P2CCD board before capacitors change.

other boards

and these are our CPU boards with 4.2 and 4.1 ROM version

SaabFAN · « **Reply #128 on:** November 30, 2017, 10:47:42 pm »

Do you have another Oscilloscope?
If so, check the PowerSupply-Rails on the large bottom board. I guess that there's quite a bit of noise there.

The next place would be to check the signal that arrives on the ADC-Board.

bd139 · « **Reply #129 on:** November 30, 2017, 11:01:49 pm »

I think all the problems I had with my PM3315 were knackered Philips Axial caps on the main board. If you don't have a scope but do have a reasonable DMM you can trace them out on low ohms range - they usually go short.

wazoo · « **Reply #130 on:** December 01, 2017, 08:48:20 pm »

yes SaabFAN , I have a rigol 1054z like you

my brother is checking power supply capacitors
step by step we are changing all capacitors.

SaabFAN · « **Reply #131 on:** December 02, 2017, 12:45:44 am »

Also check the Clock-Lines to the P2CCDs and all the other control-signals. The CCDs are analog devices and it is quite likely they're sensitive to the clock-level.

The clock is generated on the driver-board that sits on top of the CRT. The one hidden beneath the shielding. That board also contains a lot of SMD capacitors that need to be replaced.

SaabFAN · « **Reply #132 on:** December 21, 2017, 02:06:49 am »

Now that I also have a PM3323, I have started a new thread specifically for the problems of the PM3323: https://www.eevblog.com/forum/repair/repairing-the-aquisition-system-of-the-philips-pm3323-oscilloscope/msg1380244/#msg1380244

The reason behind that is that the PM3323 seems to be suffering from far worse capacitor-problems than the PM3320A. Most likely because the PM3323 runs a lot hotter because of the higher Clock-Frequencies.

Issues with the digital systems and the upgrade-attemps will remain here, but also apply to the PM3323, as that scope shares the majority of the components with the PM3320A.

idrenth · « **Reply #133 on:** June 29, 2020, 09:17:54 pm »

Quote from: Stonewerk on August 22, 2015, 05:56:44 am

Quote from: Stonewerk on August 22, 2015, 02:57:13 am
...
... terminal program on your PC - I use Realterm, available on SourcefForge. ...
...
....is with an HPGL viewer, such as HPGLView...
...

Sources:
Realterm: http://realterm.sourceforge.net/
HPGL_View: http://service-hpglview.web.cern.ch/service-hpglview/hpglviewer.html

I'm very pleased with the operation of both of these utilities...

Hello everyone. I just own this scope also and would like to try out the rs232-c connection. I just use the fdti232r and configure it on 5v. When I connect the TX,RX and ground so that would be on the DB25 connector pin 1 GND, 2 TXD and 3 RXD. Do I miss something? I use the same software and configure also the baud rate correct. Do I miss a step ? I send something but I don't get anything in the hyperterminal from realterm

tunk · « **Reply #134 on:** June 29, 2020, 09:38:48 pm »

It could be because it's using RS232 voltage levels, and not 5V:
https://en.wikipedia.org/wiki/RS-232#Voltage_levels

idrenth · « **Reply #135 on:** June 29, 2020, 11:06:48 pm »

That could be, but still what did they use and is the only 3 pin connection enough to work ?

SaabFAN · « **Reply #136 on:** June 30, 2020, 04:22:28 pm »

If I remember correctly, you need at least 2 handshake-Signals.

I'm not sure about the signal-levels, though. Better check it first before you fry your chip and/or a USB-Connector of your PC.

idrenth · « **Reply #137 on:** June 30, 2020, 10:55:32 pm »

I have the cable now working from db9 to a db25. I use pinout from pm8956A. Next plan is to get the FDTI232 chip working for this connection. I am also thinking about to upgrade my unit with version 5.1 it still runs on v2 Waiting for some new eproms to program them.

texaspyro · « **Reply #138 on:** July 01, 2020, 05:27:27 am »

The standard signal connection to bypass handshake signal requirements on a piece of equipment with a DB25 connector is:
4-5
6-8-20

idrenth · « **Reply #139 on:** July 01, 2020, 11:34:22 am »

So have it now working with a pl2303 chip and goes to USB connection to my laptop. That works fine. Now was wondering you guys have maybe some nice things to use on the rs232 connection prog's to read out more out the PM3320A device? I am also thinking about the gpib connection to the readout. any benefits between the both ? any good suggestion software to use with this scope?

idrenth · « **Reply #140 on:** July 01, 2020, 12:36:39 pm »

You mean to jumper them on a db25 so link pin4 and 5 also link pin 6 with 8 and 20? is that way I only need the rx tx and gnd connection still? because I have it now in an otherway to get it working from db25 to db9.

SaabFAN · « **Reply #141 on:** July 01, 2020, 03:42:48 pm »

GPIB has the advantage of a much higher bandwidth. About 1 MBit/s I think.
Unfortunately, the GPIB Firmware inside the scope doesn't know some standard commands (like ID?) that were introduced around 1990, so several programs might have difficulties talking to the scope. So it might be a bit of a hassle to set up some modern software.

So far I've been using only a terminal program and HPGLviewer to capture and display data from the scope. Don't know about any other programs that can use the serial port. There seem to be some programs out there in the wild that can make use of the GPIB-Interface, though. Haven't checked that out though, since so far I didn't want to spend money on an expensive GPIB to USB Interface and didn't find the time to make one with an Arduino

idrenth · « **Reply #142 on:** July 01, 2020, 07:08:32 pm »

I have gpib pci card from national instruments and a usb own made gpib cable. So far I know there where 2 commands for ID? and IDN? I remember that because there was HPIB and GPIB. Also I remember that fluke use IDN? and most HP and other use ID?. So make sense if you need for the philips/fluke PM3320A also maybe the command IDN? I have to check that later out. Maybe that somebody already had it working. So I could look how to use it or maybe can use it instead of redo all the research. Maybe I can look it up in the manual from the PM8956A for the command list.

idrenth · « **Reply #143 on:** July 01, 2020, 11:49:45 pm »

In the manual from the 8956a is also written that it has a command known as the "IDT ?" so the scope has it also. With the GPIB I had response "IDT PM3320A.V02 /PM8956A.V01"

When I send the command over rs232 I got a response back "IDT PM3320A.V02 /PM8956A.V01"

SaabFAN · « **Reply #144 on:** July 02, 2020, 06:55:37 am »

Right, I remember now. That command works, but when I tried out the National Instruments-Software for a GPIB-Card I had bought for my Retro-PC, the software kept using the "IDN ?" command to search for devices, which resulted in Bus Timeout-Errors.

That's what I meant by my earlier comment. More modern software seems to assume devices understand the 1990 GPIB Standard Commands and when they don't, it leads to issues

Would be interesting to see if the software for the PM8956A-Card (the one that makes RS-232 and GPIB possible) can be updated to understand newer commands. Best case would be if the engineers at Philips used some sort of look up-table that can be easily identified by looking at the firmware in a Hex-Editor or with a Disassembler.

idrenth · « **Reply #145 on:** July 02, 2020, 09:13:45 am »

Quote from: SaabFAN on July 02, 2020, 06:55:37 am

Right, I remember now. That command works, but when I tried out the National Instruments-Software for a GPIB-Card I had bought for my Retro-PC, the software kept using the "IDN ?" command to search for devices, which resulted in Bus Timeout-Errors.

That's what I meant by my earlier comment. More modern software seems to assume devices understand the 1990 GPIB Standard Commands and when they don't, it leads to issues

That wouldn't be a problem you can change the string that send or replace that string in the blocks with your own made block.

Quote from: SaabFAN on July 02, 2020, 06:55:37 am

Would be interesting to see if the software for the PM8956A-Card (the one that makes RS-232 and GPIB possible) can be updated to understand newer commands. Best case would be if the engineers at Philips used some sort of look up-table that can be easily identified by looking at the firmware in a Hex-Editor or with a Disassembler.

It doesn't really matter to have disassembler, because the commands are written in the manual should be possible to change the command in the newer programs config files. I thought it called file *.ieee488. Philips had a program to communicated with it called "BINPROG" I don't own this program but also not sure if that would be better now this days.

idrenth · « **Reply #146 on:** July 11, 2020, 09:30:01 pm »

So, I also upgrade my scope with V5.1 cpu board. I was wondering does someone have a picture where the FFT function should be? because I already had the Histogram option or is that the same for your FFT function? Maybe picture possible or a step by step to see?

idrenth · « **Reply #147 on:** July 11, 2020, 10:01:17 pm »

Here some captures from it.

SaabFAN · « **Reply #148 on:** July 12, 2020, 07:04:37 am »

The FFT-Function should be right underneath the Histogram-Function in the menu.

It is enabled somewhere else in the instrument, but I haven't been able to figure out where yet.

idrenth · « **Reply #149 on:** July 12, 2020, 09:22:18 am »

Does someone have an image how it looks? because I only see talking about it but no one has an image or a video to get there ? can somebody provide that with the 3320A? if so I will think it would be at ccd part. that is doing the calc part. but I first want to see who has it with some images and if possible a video would be nice how to know where to see the FFT.


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Repairing and Upgrading an old PM3320A - Including adding FFT! (Read 50130 times)

Share me