Elektronika MK-52: The MK-61 big brother?

[jebem]

While the Elektronika MK-61 and MK-52 presents a common set of features, the MK-52 adds constant memory and external IO ports.

Usually constant memory is implemented using static RAM with a power backup mechanism.
Not in the MK-52: It uses a writable ROM to allow users to save their programs on demand.

I got this MK-52 at a fair price, and despite the seller claiming for a good working condition, this is not the case.
Also the case was opened, as the back screw seal is missing.
So this will be one of my next restore projects.

It powers ON, but the display flickers, then shows all the dots like the batteries were depleted (using fresh alkaline AA batteries).
However I managed to do some math operations and it did OK on that, meaning that at least the logic circuits are in good shape.

Somehow I was expecting this behavior. These machines are old (vintage) and were stored for many years in a shelf, usually in harsh environmental conditions.
Electrolytic capacitors in the DC-DC converter power supply are usually the first components to fail under these conditions, specially when reapplying energy on them.
These caps where made in 1984, while the calculator was made in 1986.
I had no time and the mood to start diagnosing the calculator, but my guess is that these caps are dry and have lost the capacitance long ago.

Surprisingly, the internals looks to be in the original condition (untouched solder joints and components).
The keyboard foam, used to keep the hard plastic keys in place without rocking too much, is sticky and needs a urgent replacement, though.

So, I need to check and fix the power supply first, then I will see if it can write to the KP1601PP1 EEPROM as well.

First set of pictures:

[jebem]

1) Information on the DC-DC converter internal power supply:
The MK-52 takes the battery supply to generate a number of Negative output supply voltages for the electronics.
The battery Positive terminal is used as Ground as well.

Common:
+BATT/GND (pin9) -> Gnd/+5V to IC's like EEPROM KP1601PP1 Ucc2 pin 24.

Input:
-V1 -BATT (pin7)

Output:
-V2 -15V (pin6) -> IC logic circuits
-V3 -27V (pin5) -> to the VFD display
-V4 -17V (pin8) -> to the -12V regulator for Ucc1 pin 12 and Gnd pin 4 of EEPROM KP1601PP1
-V5 -40V (pin4) -> to the -33V regulator for programming voltage Upr pin 15 of EEPROM KP1601PP1
~V6: 2.4VAC (pin 1 and 3) -> to the VFD display filament [Edit]

Transistors:
VT1: кт361г (KT361G) - PNP Si 250MHz, hfe>50, 35V, 0.05A, 0.15W
VT2: кт315г (KT315G) - NPN Si 250MHz, hfe>50, 35V, 0.1A, 0.15W
VT3: кт626A* (KT626A) - PNP Si, 75MHz, hfe>40, 45V, 1.5A, 6.5W
VT4: кт315г (KT315G) - NPN Si 250MHz, hfe>50, 35V, 0.1A, 0.15W
* Or кт814б (KT814B) - PNP Si, 3MHz, hfe>60, 50V, 3A, 10W

DC-DC Converter diagram:



2) Systemboard main components:

IC's:
D1: К745ИК1302-2 (K745IK1302-2) - Control Unit for Processor, Display and Keyboard;
D2: К745ИК1303-2 (K745IK1303-2) - Arithmetic Processor;
D3: К745ИР2-2 (K745IR2-2) - Memory Registers;
D4: К745ИР2-2 (K745IR2-2) - Memory Registers;
D5: К745ГФ3-2 (K745GF3-2) - Clock Generator.
D6: К745ИК1306-2 (K745IK1306-2) - Additional MK-61 Processor features not present on the B3-34
D7: К745ИК1801-2 (K745IK1801-1) - RAM memory System bus I/O.
D8: КР1601РР1 (KR1601RR1) - EEPROM 1024 x 4 (Eq. to ER2401)
D9: К561ЛА7 (K561LA7) - 4x2 input NAND (like CD4011AE)

Updated: Additional notes on the IC's, by Sergei Frolov:
K745IK130x are same CPU with different firmware.
K745 is a package-less version of K145 ICs.
K745IK1302 is a main CPU (control processor). It organized initialization of serial data ring, keyboard and display controller and perform some operations.
K745ИК1801 is a external I/O controller (Internal EEPROM and External ROM module in MK-52)
K745ИР2 both serial RAM registers.


VFD display:
VL1: ИЛЦ2-12/8Л (ILC2-12/8L or ILZ2-12/8) - Vacuum Fluorescent Display. Filament voltage: 2.15 -2.9 V. Filament current: 20 - 24mA.
Brightness: 700 cd/m^2. Number of digits: 12 (8+2 digit sign).
Size of the digits: 2.4 x 4.3 mm. Color: Green.
Lifetime: more than 10000 hours.

Transistors:
VT1: кт315г (KT315G) - NPN Si 250MHz, hfe>50, 35V, 0.1A, 0.15W
VT2: кт361г (KT361G) - PNP Si 250MHz, hfe>50, 35V, 0.05A, 0.15W
VT3: кт361б (KT361B) - PNP Si 250MHz, hfe>50, 20V, 0.05A, 0.15W
VT4: кт315г (KT315G) - NPN Si 250MHz, hfe>50, 35V, 0.1A, 0.15W



Second set of pictures:

[jebem]

DC-DC converter power supply troubleshooting:

As I had a couple of hours to kill before my tonight flight to Cuiabá (Brazil) for another work week, I have removed the four electrolytic capacitors to test.

As I suspected, two of them are dried and have lost the capacitance (see measurements in the pictures below).

Two capacitors are from one Russian brand, and the other two are from another Russian brand.
Interestingly, the failed capacitors are from the same brand, despite having being manufactured after the other two:

- Two 5uF / 50V, Ref. K50-6, made in 0384 (March-1984), are OK.
- One 20uF / 25V, Ref. K50-16, made in 8509 (1985-Sept), is dead.
- One 2uF / 25V, Ref. K50-16, made in 8603 (1986-March), is dead.

[jebem]

Repairing the DC-DC converter power supply:


I couldn't wait, could I?
My flight is 5 hours away and I'm still at home doing this nice hobby of ours... I guess it is time to start packing now.

Replacing the defective capacitors:
So, I just took the capacitors that I had at hand, and have selected these Jackon brand Low ESR 105c high temperature electrolytic capacitors able to cope with harsh environments and working nicely in switching power supplies as this one.
It should be more than enough for the job.

One word of caution for those who are whiling to do this repair as a diy hobby:
The PCB is double face and fragile, so use a small thin tip but potent iron solder to melt the solder joints in order to remove the components without bringing the PCB copper traces attached to the component leads.

I decided to replace all four capacitors with new ones, despite the fact that the two original 5uF are good.

After connecting a 6VDC power source, have powered ON the MK-52 and voila, it works!





Voltage adjust procedure:
However the output voltages were off the mark, so I had to adjust it to the correct values.
I have connected a Voltmeter between the +Vbatt and pin 6 of the DC-DC converter and adjusted the potentiometer to read -15VDC.
Then check the other DC output voltages - they should be correct within +/-5%.

The current consumption is about 95mA, which is average for this kind of machines.





Checking the DC-DC oscillator waveform:

Just for fun of it, I checked the switching oscillator wave shape.
My frequency meter is dead, waiting for a new set of NiCd batteries (Fluke 1920A), so I have used my vintage Centrad scope to estimate the frequency as well.

The period is about 14uS (71KHz) at 15Vpp, which is not far away of what the Elektronika is showing in the MK-61 schematics (I couldn't find this information for the MK-52).





Assembling all together and initial RUN mode tests.
It looks good.
(I'm leaving the EEPROM testing for later):

[jebem]

Final test: Check the EEPROM constant memory

What the heck, I had just time enough to do this verification test before preparing myself to travel.

I have to thank again to Sergei Frolov for publishing so much details on these Russian calculators.

The next procedure was prepared by Sergei.
I followed it by the letter, and my calculator works nicely, saving the program area contents to EEPROM!!!

To ensure that your MK-52 has workable EEPROM, you may proceed
following test:

Erasing EEPROM:
1. Place 'C 3 CЧ' switcher -> 'C' position (clear EEPROM)
2. Place 'Д П' (D P) switcher -> 'П' (P) position (mean "Program
area")
3. Turn on calculator or press [Cx] key.
4. Enter 1000024 (first 24 cells to erase)
5. Press [A^] (A up arrow)
6. Press [^v] (up/down arrows) (clear EEPROM)

Filling EEPROM
7. Press [B/0] (program counter = 00)
8. [F] [ПРГ] (PRG) (enter program mode)
9. Press [1][2][3][4][5][6] (or something similar keys)
10. [F] [ABT] (return to manual calculations).
11. Place 'C 3 CЧ' switcher -> '3' position (write to EEPROM)
12. Press [^v] (up/down arrows) (store program to EEPROM)
13. Place 'C 3 CЧ' switcher -> 'CЧ' position (read from EEPROM)
14. Turn OFF calculator, wait for 30 secs.

Checking EEPROM
15. Turn ON calculator
16. Enter 1000024 (first 24 cells to reading)
17. Press [A^] (A up arrow) (store command)
18. Press [^v] (up/down arrow) (read EEPROM)
19. [F] [ПРГ] (PRG) (enter the program mode)
20. Press [ШГ->] (step forward) six times and check for codes [01]
[02] ... [06], which was entered in step 9.

[Gene]

No one has commented yet, but don't take that as no one being interested.

Thanks for this trip down memory lane. I have the MK52 and MK61 and this is fascinating to me.

Way to go. Safe Trip!

[Didier Lachieze]

I don't have a MK-52 nor MK-61, but I enjoy reading your posts about these machines!
I may end up getting one of them but currently at home the WAF for any new old calculator is pretty low...

[jebem]

(10-28-2014 06:56 PM)Gene Wrote:  No one has commented yet, but don't take that as no one being interested.

Thanks for this trip down memory lane. I have the MK52 and MK61 and this is fascinating to me.

Way to go. Safe Trip!

Thanks, Gene!

My goal on the MK-61 and MK-52 was to have the hardware details, repairing, and usage documented it in one single place for easier access.
After this, they will join my modest collection along with the others (HP, TI and CASIO).

The flight to Sao Paulo, Brazil was normal as usual (TAP is a safe company to fly, by the way).
Then I took another domestic flight to Cuiabá, Mato Grosso, Brazil.
Cuiabá city temperature is extremely high today!
At noon I had to go out and walk to the shopping center to have lunch, and one could fry an egg in the street walk!
Still, the landscape shows a fantastic green color from the vegetation.

[jebem]

(10-28-2014 08:12 PM)Didier Lachieze Wrote:  I don't have a MK-52 nor MK-61, but I enjoy reading your posts about these machines!
I may end up getting one of them but currently at home the WAF for any new old calculator is pretty low...

I have the same WAF constrain. Otherwise, I can guarantee that by now my apartment would be like a museum in every single room!

[Francois Lanciault]

I was hoping this could be my problem too... But after removing the four capacitors, they all test good. I will put everything together and check the voltage.

François

(10-28-2014 01:21 PM)jebem Wrote:  DC-DC converter power supply troubleshooting:

As I had a couple of hours to kill before my tonight flight to Cuiabá (Brazil) for another work week, I have removed the four electrolytic capacitors to test.

As I suspected, two of them are dried and have lost the capacitance (see measurements in the pictures below).

Two capacitors are from one Russian brand, and the other two are from another Russian brand.
Interestingly, the failed capacitors are from the same brand, despite having being manufactured after the other two:

- Two 5uF / 50V, Ref. K50-6, made in 0384 (March-1984), are OK.
- One 20uF / 25V, Ref. K50-16, made in 8509 (1985-Sept), is dead.
- One 2uF / 25V, Ref. K50-16, made in 8603 (1986-March), is dead.

[Sergei Frolov]

(10-27-2014 06:06 PM)jebem Wrote:  [/b]D1: К745ИК1302-2 (K745IK1302-2) - Control Unit for Processor, Display and Keyboard;
D2: К745ИК1303-2 (K745IK1303-2) - Arithmetic Processor;
D3: К745ИР2-2 (K745IR2-2) - Memory Registers;
D4: К745ИР2-2 (K745IR2-2) - Memory Registers;
D5: К745ГФ3-2 (K745GF3-2) - Clock Generator.
D6: К745ИК1306-2 (K745IK1306-2) - Additional MK-61 Processor features not present on the B3-34
D7: К745ИК1801-2 (K745IK1801-1) - RAM memory System bus I/O.
D8: КР1601РР1 (KP1601RR1) - EEPROM 1024 x 4 (Eq. to ER2401)
D9: К561ЛА7 (K561LA7) - 4x2 input NAND (like CD4011AE)

Hi!
I have a little notes for ICs:
K745IK130x are same CPU with different firmware.
K745 is a package-less version of K145 ICs
K745IK1302 is a main CPU (control processor). It organized initialization of serial data ring, keyboard and display controller and perform some operations.
K745ИК1801 is a external I/O controller (Internal EEPROM and External ROM module in MK-52)
K745ИР2 both serial RAM registers.

[jebem]

(11-15-2014 11:38 PM)Francois Lanciault Wrote:  I was hoping this could be my problem too... But after removing the four capacitors, they all test good. I will put everything together and check the voltage.
François

Hi, Francois,
I started by checking the output voltages first, and it were wrong and with a lot of ripple noise (to check this you need a oscilloscope besides the multimeter).
In my case, the problem was on the defective capacitors.

Sometimes the capacitors measures well on the capacitance meter, but this do not mean it are good for use.
The reasoning here is that the vast majority of capacitance meters are not able to check the capacitors at the required operating frequency of our application.
This is a switching DC-DC converter operating at a very high frequency, so the capacitors may be good at low frequencies like 1 or 2KHz, but it will fail at the required operating frequencies used here (about 70KHz).

So, assuming you have wrong output voltages, and when in doubt, replace the capacitors.
If it do not fix the problem, continue to check the remaining components in the DC-DC converter, but I would recommend you to acquire an oscilloscope to do the checking.
It is a nice investment for the future and you can find plenty of them at eBay for low cost, specially the used models from the 70's and 80's.

[jebem]

(11-16-2014 09:15 AM)Sergei Frolov Wrote:  Hi!
I have a little notes for ICs (...)

Привет / Hello, Sergei!
Спасибо большое / Thank you for your contribution, I will correct my posts above with your information.

And welcome to this nice HP Museum!

[jebem]

How to use the EEPROM to save programs and data.


Looking to the published examples in the Internet and wanting to know a little bit more about the memory structure details, I had a look into the MK-52 Instruction guide in Russian.
It turns out that the way the information is presented is on the dark side and with one or another imprecision.

The attached draft document tries to summarize my findings so far.

  MK-52_EEPROM_usage.pdf (Size: 464.53 KB / Downloads: 125)

[Geoff Quickfall]

Excellent work Jebem!

Have been programming and was only having difficulty in storing multiple programs in the eeeprom.

You cleared that up with the requirement for address start to be multiples of 16; that is 16, 32, 48.

Also I could not find a label key, so specifying to an address line helps with conditionals and branching.

Will play some more and let you know if I find anything new. It sure is SLOW, but fun to watch as it runs.

Geoff

[Francois Lanciault]

I had no time to try to fix my MK-52, so instead I ordered an other one! Thank for this new set of instructions, it will be of great help I am sure.

François

[jebem]

(01-06-2015 12:32 AM)Geoff Quickfall Wrote:  Have been programming and was only having difficulty in storing multiple programs in the eeeprom.

You cleared that up with the requirement for address start to be multiples of 16; that is 16, 32, 48.

Will play some more and let you know if I find anything new. It sure is SLOW, but fun to watch as it runs.

Geoff

Hi, Geoff!
You are welcome!

But please keep in mind those MK-52 notes are not validated, as it are just a compilation of several articles found in the Net plus some sentences extracted from the Russian guides.
Apparently the calculator works in so many ways that it is a mighty work to document in full.
Perhaps that's why many say that it have many "undocumented features".

Yes the slowness is just amazing!
It gives me the feeling that the machine is really busy doing a hard work!
Well, what can we expect from a +/- 400 KHz clock speed after all?

(01-06-2015 05:49 AM)Francois Lanciault Wrote:  I had no time to try to fix my MK-52, so instead I ordered an other one! Thank for this new set of instructions, it will be of great help I am sure.

François

You are welcome, Francois.

I'm curious about your MK-52 issue: Are the DC-DC convert output power supply voltages correct?

[jhl]

Thank you for all the useful information! I'm waiting on a dead мк-52 -- which will hopefully just be the PSU too...

I got curious and had a dig through the schematics looking at the IO ports, thinking it might be entertaining to hook something up to the poor thing. XS3 looks like a straight-up read only ROM bus (assuming the вио line on pin 9 is to disable the built-in EEPROM - does anyone know what that's short for?)

I had thought the other port was for controlling external gadgets in a similar fashion to the мк-64. Indeed, the serial data loop runs to that port. But it has three lines to set clear/write mode, and it's also got a line to all the serial data devices *except* the EEPROM controller via си (sync line) - it looks like that port is designed to talk to the EEPROM controller directly and bypass the rest; it can even feed power in to the +5 input, suggesting it's a factory test/programming port.

Does anyone know if there were any IO peripherals for the -52? I can't find anything in English. As it stands it looks like output is only possible by sniffing the serial bus or sequencing reads to the ROM port.

[Sergei Frolov]

(02-10-2015 10:45 AM)jhl Wrote:  Does anyone know if there were any IO peripherals for the -52? I can't find anything in English. As it stands it looks like output is only possible by sniffing the serial bus or sequencing reads to the ROM port.

We have only around 4 external ROM modules with programs (called BRP)
Another IO devices was not in serial production.

[d b]

Sergei; What functions do the 4 known ROMs preform? Were any of yours applicable to the Soviet space program?