HP 9403A first look inside

So .... time to look inside the HP 9403A device! I didn't do it in Friedrichshafen last June because the screwdriver I had was not the right fit and I was afraid to destroy the head. With the right tool it turned out that screws were secured with a dab of glue, then off they came easily. Once removed the thick aluminum top lid I was enchanted by the shining boards, gold plated ICs, perfectly grouped and routed wires as if it was build yesterday, while the thing is about 45 years old and has certainly passed few owners.

The absence of dust inside is easily explained: there are no holes, so no way out for the air and no way in for contaminants.

There is a lot of room inside, a power supply, an optional PSU, no microprocessor whatsoever at first glance. Board are mounted with servicing in mind, or mostly quick replacing. 90% of integrated circuits are branded HP and carry their specific mysterious 1820-xxxx part number. Unfortunately Nixies are hidden behind a part of the chassis and I cannot see their model. Nevermind. Since everything looks clean I decide to power it up.

As a precaution I screwed back the lid in place and brought the 9403A on the balcony so that exploding stuff would stay inside and any smoke would not pollute home. Well, I was not surprised that nothing bad happened and Nixies lit up:

1-2-3 ... 13 tubes! Neat. They have gas. Further experiments will tell if they have cathode poisoning or broken digits.

Any interaction with the keypad or front panel keys do not lead to any change on the displayed value. Given that there is no intelligent part inside I was expecting it too. Next step is accessing the Nixies to see their model.

Seven clocks and counting

As of December 2018 I have built seven clocks:

1. Binixie with two Nixies B5092
2. Adapter chinese LED clock kit to 4x IV-6
3. HP LED on my bedside
4. Repurposed Olivetti Logos 262PD desk calculator
5. Repurposed dual-needle meter
6. Self-designed 4x ZM1332/NL5870 Nixie
7. Single IN-14 clock with AC timebase

Four of them have gotten a proper case, while two are a repurposed commercial device that already had a box.

I would like to repurpose an old 3.5 digit multimeter: will it be #8?!

Sending fragments of data to a microcontroller

Alright. I have (autogenerated) a new problem to solve: I want to be able to send few bytes of information to a microcontroller installed into a box without using a physical input device (a.k.a. "button"). This operation has to happen more than rarely, so I would like to avoid throwing in some hardware that potentially will never be used or would lock me in to an external technology. At one-off production run it is purely a stylistic affair, not a cost-cutting measure.

Specifically, the device is a digital clock and I want to be able to set the time without extra drilling the box and installing a button. So "wirelessly". There are two holes, one for DC input socket and one for the Nixie tube protruding out of the cover. The display acts as a feedback to the user, of course.

My night-long brainstorming came up with these options:

• Plain old button(s) requires an extra hole.
• WiFi or Bluetooth are too power hungry, too costly, too technically advanced because they require a software counterpart.
• A reed switch, but the box is quite thick and it requires a powerful magnet that might not be available when needed.
• An IR receiver under the Nixie and a TV remote as remote controller that might not be available when needed.
• A microphone to receive AFSK or similar, but requires a software counterpart.
• An open capacitor to take advantage of hand proximity effect.
• A LDR under the Nixie tube and operator's hand to command through shade/light.
• A tilt sensor, not the cheapest way.

The following table summarises my options:

Type	Hole?	Extra tech?
Button	Y	N
WiFi / BT	N	Y
Reed switch	N	Y
IR receiver	N	Y
Microphone	N	Y
Open capacitor	N	N
LDR	N	N
Tilt sensor	N	N

I like the last two the most. Both LDR and tilt sensor are hand-operated and allow for an easy repurpose too: at powerup they serve as input devices to set the time, during operation they can provide information to control brightness.

I need to verify whether the LDR under the Nixie receives enough light to understand "light vs shade" situations. Otherwise I will order a tilt sensor.

Readers, please leave a comment if you have further interfacing ideas!

Reviving a ceiling lamp from CFL to LED

We've had at home a ceiling lamp for about 10 years, and for the last 4 years it has misbehaved. It contains two 55W CFL tubes with an external HV power supply by OSRAM (Quicktronic Professional QTP-DL 2x55 GII).

The defect was that it would switch off after about 15 minutes of operation and it would restart only after a rather long while. In any case it was not usable at all.

The original configuration.

I tried replacing the gas-filled elements without success, so the culprit must have been the embedded power supply. Since I am still developing my spider superpowers, working upside down on the ceiling is not my thing so I pulled down the lamp on the table. Still about 15 minutes of operation even without the front glass (in case it was a heat issue).

I took apart the power supply case and looked for the obvious, that obviously wasn't that obvious. I simply reconnected everything with the power supply out of its case and it would run for hours. HA! Maybe a cold solder joint failed with heat cycles over time and moving the circuit around sort-of fixed it.

I took my chances and put everything into the metal case, and still working. So I secured the case back to the lamp and ... 5 seconds of operation!

Enough. Either I got another identical power supply or I had to find something else. My readers probably already guessed which one I picked! "Something else", of course.

The lamp base is metallic and hidden against the ceiling so it can be drilled as needed to accomodate a different setup.

Few searches later I learned that there are LED replacements for neon tubes, and they include the power supply too. I chose two circular replacements whose total equivalent power would get close to the original 110W.

Of course I had a look inside the power supply, that is very simple since it has to light up a series of LEDs:

The LED strip power supply.

And the final look, up on the ceiling before placing the cover glass:

Final config. Notice something "light"?

Now the two light sources are independent, so if one fails the lamp will still work at 50% luminosity. If you look closely at the last picture you will see LEDs are "on", dim, but "on". Even with a single wire attached, the culprit is probably some leak of the long driving wire towards ground. I couldn't get rid of it by swapping the two cables, so I consider it as a night light undocumented feature.

We don't use that lamp very often, so hopefully it will last another 10 years!

Elka 135, calculator from Bulgaria

Meet the Elka 135 scientific calculator! I bought this marvel in FN 2018 fair. Batteries had leaked inside, so this was the fault to fix (yet again).

There are few pages online that describe this calculator made in Bulgaria. I like the 8-digit LED display. It can be powered from an external 6V DC source through a 3,5mm audio jack (!!).

Using a variable voltage power supply I could confirm that it works down to 4.5V and a small "L" appears in the leftmost position: this means I can use this calculator with USB power, without fully restoring the battery contacts.

Few keys are bouncy resulting in double-triple press, but I am not sure whether they were like that in the original, non-damaged calculator.

More complex operations take some time, and the left position in the display keeps updating until the result is ready.

This device is quite power hungry, because of the LED display, but modern batteries (or USB power) can handle it.

Industria Macchine Elettroniche IME-141 - eeeeeerror

I let my daughter play with IME-141 calculator. Since a calculator is boring if you don't have maths to do, she pressed few buttons at the same time and this is the result:

It never occurred to me I could to that "test". Now I try it on every calculator just for the fun of it.

And for the curious, a close-up of display VFD tubes. I really like the shape of segments. The vertical filament is visible too.

FM radio for the bathroom - followup

This is a follow-up to my previous post about an FM radio for the bathroom morning routine.

The good old Samsung Galaxy S (i9000) does a wonderful job. Its stand-by time is more than a week even if we leave the GSM_RF part powered all the time (otherwise we would need to switch it on every time we need the FM radio). The loudspeaker volume and quality are very pleasant.

I even found a smaller 3.5mm jack to sacrifice and now it looks even better without effect on the received signal.

Long live radio!

Case closed :)

Rethinking my mobile setup

Alright, the shortened UHF-only Nagoya UT-106 is back in service.

I added a piece of RG-58 with PL-259 so that I could measure SWR. But just the fact of terminating the coax with a PL-259 made me think of a way to fit the QYT KT-8900 mobile in the new car in place of a handheld.

While the large glass roof drastically reduces the flat metallic area for a magmount antenna, it does not block RF either. Since the only place for a small magmount is above driver's head, the antenna cable would be in the way of the passenger because I don't leave the antenna outside when the car is unattended, go figure. But there is plenty of space in the back of the car. Yes, inside, but with all those transparencies (be it glass or plastic), RF-wise it is almost like being outside.

The plan is to place the magmount base on a metallic surface and let it rest on the parcel shelf. At home I found an L-shaped 40 cm long iron piece that can be forced between the back seat and the shelf. SWR of this combo is 1.5:1 on the FT-817. So far so good.

Will it work?

Nagoya UT-106 intermittent signal

Now that I was enjoying trouble-free mobile HAMming, my modified Nagoya UT-106 developed a fault causing intermittent reception. Moving the coax near the RTX would restore/loose the signal, so probably something has broken in the coax. Since the handheld lays on the passenger's seat, holding the microphone causes it to move around and put some stress on the cable.

I will cut out the broken part and try to resolder the SMA connector. Before that I will fit a PL-259 so that I can try the whole setup with a more powerful RTX as well as measure SWR.

FM radio for the morning routine: September 2018 attempt

Following my previous post on two attempts at joining the morning news and the morning bathroom routine, in both cases something is too wrong. Memory lost and instable signal are no-go.

Initially I admitted my failure to provide a suitable device and let the family member head to the shop to look for a replacement. Then an idea strike: why not use the FM radio into that old Samsung Galaxy S phone that was gathering dust? Just fit a 3.5mm jack (with enough wire) to behave as an antenna and set the output through the internal loudspeaker.

Bingo! While an overkill, it is battery powered, rechargeable and does not loose memory when the battery goes flat (but you can monitor it easily). It has more than enough audio power to fill the bathroom and costs nothing. Last but not least there is a self power-off setting at 30 minutes! The only drawback is that the RF-GSM part has to be ON to receive FM radio. But that works even without a SIM card inserted.

As shown in the picture, if the FM station is close/strong enough a 3.5mm jack alone acts as a perfect antenna (picture shows a 5mm to 3.5mm adapter). Probably it also helps not to overload the input stages resulting in better reception!

Let's see how long this solution lasts.

Note. I know I could have used an Android app to listen to the Internet streaming transmission. But who knows if there is still a working app around for Android 4.x? And how it would compare with power consumption? And the current over-the-air solution works even during power outages.

FM radio for the morning routine

The "old-style" replacement RX-16F

I was appointed an apparently simple task from a family member: have an FM radio in the bathroom to listen to the morning news. It is basically a single frequency FM receiver outputting audio through a loudspeaker. Possibly as small as possible and requiring as little maintenance as possible. Something: "flip a switch and listen to the news."

If we were in the 1980's I would buy an analog tuned pocket radio powered with two AA batteries. Unfortunately they don't seem to be produced anymore, replaced with energy-hungry digitally tuned radios (with a varying degree of user friendliness user interface). Used devices have all sort of problems.

4 years ago I bought a small cube powered with a rechargeable Lithium cell (Kevler KP-520PLL). It would receive FM just fine with a big problem: current consumption. Even when powered off it would drain the battery flat in a week (2mA when off, 48mA when on). What is worse is that it looses memory of the last tuned station when the battery dies and the single-button tuning system makes it really annoying to reset way too often.

Then in Marzaglia ham flea market in May 2018 I picked a 1970's analog tuned FM radio that sounded good in the field, marked RX-16F. This one, while lasting forever on a couple of modern AA batteries, suffers a long forgotten (and solved) problem: the hand proximity effect! If you hold the receiver in your hand it would behave perfectly, but looses the tuning and strength (no AGC!) when left on the table. It's fine for listening to the football match while walking downtown, but not practical for the bathroom morning routine.

I need a better solution.

Back of the FM radio.

RX-16F circuit.

Flea markets in Croatia

I think Croatia's population of 4.1 million does not justify the presence of thrift stores. In fact I could not locate one that would sell electronics and related items. The key to success in this case is to know the right name in the local language. That's how I discovered periodic flea markets: "sajam".

While I have always known of Hrelic market in Zagreb, and never managed to pay a visit because I am never there on Sunday morning, searching for "sajam" on an Internet search engine brought up other events.

Like in Sesvete (East of Zagreb) every Thursday morning. Or in Samobor (N-W of Zagreb) every Saturday morning. And in Split every Sunday morning.

All of them open quite early, so you can get there at 6-7 A.M. and be back to your usual business before anyone notices.

I could visit Sesvete sajam on 2018-08-23. No pictures, sorry. It is a mix of farmers' (wholesale, sort-of), textile and flea market. Given the holiday period my impression was that the flea area was not too crowded. Best to get there by car. Electronically speaking, there were few sellers with this kind of items including one stand full of 1980's radio receivers. As usual you need your usual luck and trained eye to spot something interesting. In Sesvete sajam there are stands serving warm food, so if your bargaining takes ages you can have lunch break as well.

If you know other periodic "sajam" around Croatia please leave a comment!

Nagoya UT-106 modified for UHF only

Over the past days I could spend few minutes with the UT-106, study the problem and solutions.

1. Magnet.

The magnet is stuck to a metallic spacer that is supposed to be glued to the antenna base. I re-glued it and reapplied the adhesive paper: it seems it holds better now.

2. Length.

Some reviewers report that the UT-106 is not cut for HAM radio bands, but resonates higher (so it is too short). I don't have a proper connector adapter to measure SWR but I trusted those reviews and applied what I originally wanted to do: cut it to a 1/4 wave on UHF just below the upper coil. The lower coil is still there, but the resulting antenna is stiffer and does not wobble freely, which means it does not pull the base all the time. As for SWR, 3 metres of RG-174 cable introduces enough losses that the transceiver won't notice an antenna mismatch.

3. Testing.

First on-air test was positive as I could hit the repeater from the underground garage, which is not possible with the UV-82 stock antenna.

Second test was on the road. I could see the antenna through the glass roof and it is stable even on bumpy asphalt at 60 km/h. Signal reports from the usual friends were positive as they know how I got through every morning with the rubber duckie.


Third test at high speed. The antenna is positioned above my head and I can see it through the glass roof. My car has no flat metallic surface, so the small base does not touch 100%. At 117 km/h the antenna begins to vibrate: probably a bit faster and something bad happens. I can try to fix this with a small spacer that reduces the front-back empty space, but that's low priority considering that I rarely drive out of town and speak on the radio.

Fourth test at "DX" distance. This surprised me most. I could hit my usual repeater from far away, through a valley where I could do the same with a 20W mobile RTX and a better performing antenna (theoretically).

So after all I will keep this antenna and enjoy mobile HAMming. But I had to modify it to suit my needs and safety requirements.



Now I do want to check the SWR.

My Nagoya UT-106 review

I needed a small VHF-UHF antenna for mobile operation. I use mostly UHF, so I picked one of the cheapest antennas on the Bay with the idea of using the connector, coax, magmount while cutting the antenna to a simple 430 MHz 1/4 wave vertical.

I ordered a Nagoya UT-106.

In the past I owned a similar dualband antenna, but with just one coil, so I can do comparisons. The very first impression is that the base is too small. The  UT-106 base is as large as 2€ coin and purely circular, while the older was slightly larger and had three small protruding feet that helped stability.

A smaller base means a smaller metallic area that couples with the car body (reflector or ground plane). Also the antenna oscillation is more likely to snap everything off the car in a bump or just at highway speed.

So I tried the magnet on pure iron. Weak. Too weak. Wobbling the antenna radiator caused the whole base to jump back and forth. I will not mount this antenna on my new-ish car and drive around!

As last check I tried to lift the bottom cover that is not made of rubber. It was not hard at all as it pulled off with my thumb and no force whatsoever. The result is visible in the picture below. Horror! The magnet is sticked to the paper base and not to the conic shape! So the antenna could fly away and leave you with a black sticky paper magnetically attached to your car! Actually there are sign of glue on the base side, but the adhesive paper won!

With all these "bugs" I will not even try it on-the-air.

I know the old say "you get what you pay for". I was prepared to do some improvements, but this is probably too much to fix.

My review: don't buy it.

Data for obsolete HP 1820 ICs

You will read everywhere that in the golden age of electronics (1960's 1970's) many parts were very short lived because newer versions became available almost each month.

This means that if you are reverse-engineering a board from those years you might meet strange part numbers that do not have modern equivalent. The problem arises if there is no manual/diagram available or you have a board of unknown origin.

Today I will concentrate on HP integrated circuits from 1970 and thereabout. Just because the HP 9403B device is full of them. They were numbered 1820-.... with no resemblance to the current CMOS or TLL naming standard.

I realised that the information is available online but it is not indexed because it is contained into PDF scans of device manuals, that in those years were written in a very detailed way: they would explain in detail how it worked theoretically and practically, schematic diagram included.

So searching for the part number will not get you far. Hoping to help the community of reverse engineers and addicted to 1970 circuits, I include below the pinout of some HP 1820 integrated circuits with plain text description found in the HP 5326B manual, so that it should become indexed by search engines. Please leave a comment if you found the information you were looking for.

List of integrated circuits currently covered:
HP 1820-0054
HP 1820-0068
HP 1820-0092
HP 1820-0094
HP 1820-0102
HP 1820-0117 (also HP 1820-0119 and HP 1820-0232)
HP 1820-0142
HP 1820-0145
HP 1820-0147
HP 1820-0174
HP 1820-0198
HP 1820-0199
HP 1820-0201
HP 1820-0209
HP 1820-0212
HP 1820-0213
HP 1820-0223
HP 1820-0238
HP 1820-0253
HP 1820-0272
HP 1820-0273
HP 1820-0274
HP 1820-0275
HP 1820-0276
HP 1820-0729 (see 1820-0092)
HP 1820-0307
HP 1820-0327
HP 1820-0328
HP 1820-0558
HP 1820-0561
HP 1858-0004

There is one image file per part number. The file name is the part number.

The zip file can be downloaded here http://bit.ly/HP1820pinouts on Google Drive or here http://bit.ly/HP1820pinoutsDropbox via Dropbox (1.1 MB).

HP Nixie driver IC (HP 1820-0729)

The HP 9403A brought another set of Nixie driver chain ICs in my home. It was the HP take on the matter: custom ICs with hard to find pinouts, let alone a datasheet. Apparently with one positive side: the pinout was laid out to simplify PCB routing.

I found an HP 1820-0729 that is an updated version of HP 1820-0092. They are a sort of 74141/7441 BCD to decimal decoder and driver but obviously with different pinout.

Ahead of that there is an HP 1820-0116 that is a 4-bit latch.

HP 09403-60006 board with buffer, driver and Nixies.

I have got an idea where to look for pinouts, but it is another story.

A cheap hands-free solution: first test

Just a quick update on the hands-free solution.

The stock earpiece-microphone of the Baofeng UV-82 is quite lousy with its muffled sound and insensitive microphone. At least that's how my specimen behaves. So it is not a big deal to sacrifice it.

I removed the microphone capsule and used it to feed the audio coming out of the wireless receiver. I added a 470nF capacitor in series to block DC. Initially I chose 10nF but the resulting signal was too low (checked on the oscilloscope). Also remember that this C makes a high-pass network together with the input Resistance, so your voice could be attenuated further.

The first on-air test resulted in low audio. Then I hit twice the [+] button on the transmitter and the transmitted volume was back to normal. An unknowing HAM did not report a different sound of my voice, so the whole thing works.
VOX can work but it is insensitive and on UV-82 there is no setting for the hold time.

Next step will be to build everything into a comfortable box, add a lever PTT and possibly a way to power up the receiver together with the RTX. It has to work until I don't receive the external antenna and check if it can handle the RF power of the mobile transceiver I still own.

A cheap hands-free solution: the idea

One of the many drawbacks of driving and talking on the radio is that the microphone is usually physically wired to the transceiver. Let alone that it is illegal in many parts of the world. Very few RTX's have a hands-free solution out-of-the-box and even in 2018 too few support some form of [Bluetooth] detached earphone-microphone. Last but not least, all off-the-shelf solutions can be expensive.

Obviously I looked for a DIY alternative. To simplify things a bit I assumed that an RTX for mobile use has enough audio power in the speaker so that an earphone or an audio amplifier is not needed. This means we need to transmit audio one-way only: from the mouth to the TX.

Now our feared Chinese fellows come very handy. Head to any of their portal and look for "Wireless microphone 2.4 GHz" (they have some that operate in UHF too). Be prepared to dig through many products until you land on something advertised for touristic guides. These products have a boom mic with transmitter and a simple receiver that is supposed to plug into an amplified speaker or other form of audio amplification.

I bought the product from NEWGOOD for 19€ (in 2017). It's still a bit more expensive that my desiderata since I hoped for 10-12€, but definitely cheaper than other solutions.

Both parts have an on-off switch, a microUSB socket for charging and the transmitter has volume control buttons. They are not Bluetooth so there's no need of pairing. I have no idea whether two similar units would interfere or could be swapped, but it's not a popular product I'd say and the risk is low even if the advertised range is 50 metres!

In its intended use, so with an amplified speaker on the other side, it works right. The audio quality is also very good.

How to interface with the transceiver then? Just feed the receiver output into the microphone input through a bypass capacitor, think of a way to handle the PTT: pushbutton, lever switch, VOX, .... and hope that signal levels & impedances match!

My current mobile RTX is the Baofeng UV-82 handheld, so it will be my test platform.

HP 9403A System Control Panel

The purchase at Friedrichshafen 2018 with the best €/kg was the HP 9403A device. And I think it will have a very high fun/€ ratio too!

Let's see how it looks:

HP 9403A in FN, near the beer kiosk.

There are three display windows, a numerical keypad, other buttons, flip switches and a key operated switch. Let's not forget the two handles for carrying. On the back side there are three large barrel connectors with a high count of contacts (40 or so), AC power, fuse, power switch.

Me and the HP 9403A.

A preliminary search on the Net before the purchase did not return meaningful data. It turns out HP has "recycled" the 9403 code with inkjet printer cartridge. The challenge is set!

The seller held the device until the afternoon, when I picked it up on the way to Hall A (with new stuff) and the car.

I had to be careful not to hit someone's legs with it. I took few breaks along the way and I even sat on it. Hans G0UPL wanted to get one too because it had Nixies and all the driving circuitry. Yeah, I know, that's why I got it in the first place.

So the mysterious device came home with me, silently opening a deep dive into 1970's documentation.

Friedrichshafen 2018

Friedrichshafen 2018 Ham Radio fair is history now.

I've been there, Friday and Saturday. My 3rd time, with 2015 first and 2016 second. You will read everywhere that there was a low both on visitors and exhibitors (on the "commercial" side), and that's true. It was extraordinarily quiet on Saturday morning! On the other hand everyone was more relaxed and easier to interact with. I finally met Hans Summers G0UPL from qrp-labs.com!

What did I bring home? For the future-me:

• QCX kit for 20m, 42€
• New Nixies from elecments13, 26€
• Assorted Nixies used from LY2MQ, 10 pieces, 20€
• Ammeter with calibration sheet, CCCP-made, 5€
• Elka 135 calculator from a Polish seller with lots of vintage stuff, 3€
• 7-segments LEDs MAN4640A in brown case, 8 pieces, 0,50€
• HP 9403A Control Panel with fun discussion, the best €/kg!

I tested used Nixies with my portable HV generator. At least they are not dull at the origin.

The calculator was too cheap to be left there anyway, so I didn't try to fit my 4xAA batteries in the compartment full of dry leak.

The mysterious HP 9403A was simply too full of Nixie tubes that could be seen through the glass. It doubled as a (heavy) portable chair on the way to the car.

What I didn't buy:

• IV-17 at 3€ ea.
• Elka 101 calculator from the same Polish seller
• Other instruments with Nixie inside (quoted at 50€)
• Nice looking valves

Dot matrix LED driver chips

While searching for something else, I came across a forum where a bunch of TIL305 dot-matrix vintage LED displays would be driven by a HT1632C chip. Neat, because it would avoid the abuse of bunch of ATtiny's and allow greater flexibility. Unfortunately the usual Far East sources don't stock a breakout board with the HT1632C already soldered and without a modern LED matrix.

DIY? HT1632C comes in 48/52-pin LQFP package that is too small for my soldering tools.

Well, digging a little further I found that Holtek produces also three versions of the HT16K33 driver, and it is commonly found on breakout boards on eBay, without LEDs attached. It is controlled via I2C and it features a matrix keyboard input too, just in case. Last but not least, its made in 20-28SOP that is manageable with a fine tip soldering iron if I ever want to make a PCB (not a bad idea since wiring is going to be a mess).

I ordered a few, so that all those TIL305 can finally spring to life.

Industria Macchine Elettroniche IME-141 - power up

A fuse holder inside IME-141

Industria Macchine Elettroniche IME-141 calculator was the one in the lot of four that wouldn't power up, so it was necessary to have a look inside. The first step was to remove three screws and I got a bit worried when I found all of them to be loose. At the first peek inside I noticed the dust (meaning not much has been done inside), the date code on a capacitor telling me (19)73 and a fuse holder: that's a synonym of hope!

Hidden into the fuse holder was this transparent component, showing to the world its inner workings and an obvious gap: the fuse had blown.

Blown old fuse

Its markings suggest it was 125V 1/16A: so large and so sensitive? Perhaps it wasn't even the original one, since there is no marking on the expected value. The closest I could find at home with the same size was 5A, borrowed from another appliance. Well, better 5A and a finger on the power switch than a fully bypassed fuse, isn't it?!

"Hello 2018 World" from IME-141.

All set: calculator switch in "ON" position, finger on the external switch ("OFF"), eyes closed in case something blows up. Click! No alarming sounds, no smells. So I did open my eyes to see the squared 0 staring at me. It works!! Fun is over. HA!

Actually I need to find a safer way to power up the beast without replacing the AC socket. Or forget the historical thingie and fit a standard one.

A lot of four calculators

This acquisition has been a very lucky find: four devices in the same lot, and three of them from early 1970's! I've had to wait a couple of months before getting my hands on them, but I think it was worth it! Let me briefly introduce them and their display technology (let's not forget I got into calculators because of their displays!):

• Casio CL-200 co-branded Lagomarsino (Nixie)
• Industria Macchine Elettroniche IME-141 (Single-digit VFDs)
• Industria Macchine Elettroniche IME-401 (Panaplex)
• Canon P34-D (VFD tube)

These devices should provide me with some entertainment, beginning with finding a way to power them up! Only IME-401 has a now-standard AC socket, while larger and older machines need a plug probably called "oval". Nevertheless I have been able to power up three of them with mixed results.

Three 1970's and one 1980's desktop calculators.

The real bounty of this acquisition could be the bunch of apparently unrelated power cords, plugs and sockets that came in the box.

RD-160: lower part of the tuning range

Before returning the chassis into the Radiomarelli RD-160 case I had to fix the lower tuning range. If it mutes then something stops oscillating. So I probed the variable capacitor with an ohmmeter and found out that the plates shorted in the first third of the range.

A closer look revealed that the outermost plate, that is made of three petals, was bent. And the middle petal was bent inwards, shorting the capacitor.

Last I cleaned up the plates blowing a bit of air and passing a strip of plain paper through all blades.

RD-160: capacitors replaced

Just a quick visual update now that I have replaced the 5 capacitors. I read an advice of buying axial capacitors as replacements and now I see why.

"All American 5" circuit of RD-160, with new capacitors.

Next step is powering it up!

Inside the Radiomarelli RD-160

My journey into the wornderful world of valves (tubes) has taken me in front of a 5-tubes radio receiver from 1955: components looked different back then!

Without a wiring diagram I've had to reverse engineer the circuit starting from the tubes. Large-ish cylinders are the possibly leaky capacitors. Two of them have like "melted" their caps: I will replace all five. I am in doubt about the large dual electrolytic, and it will stay for now.

Inside RD-160 - original!

In order to reconstruct the wiring diagram I took a picture of the circuit, cut it on the computer to show only relevant parts, increased brightness and printed it out on A4 paper. This way I can add notes on paper while I understand which component is which. I wanted to print it in B/W, just borders, but it wouldn't be easy to distinguish between components on different layers.

Three out of five "suspect" capacitors have lost their markings. I will chop them off so that their leads will be used as support for replacements. The capacitance meter confirms their expected values, within 20%. Still, I don't trust them.

Now waiting for the order to arrive. I chose replacements rated for 630V and I bought packs of 10, just in case I will step into another valve radio to restore (I want one with the wooden case!).

Canon Palmtronic 8 (LD-84)

Yes, another calculator showed up at home. This time it is a pocket device called Canon Palmtronic 8, also known as LD-84. It works off 2xAA batteries and has 8 digits shown on a VFD display. It also accepts an external 3Vdc power source.

This one did not need any restoration, except for some light cleaning of the plastic case.

The rightmost digit is slightly fainter than others because obviously it has been "on" longer than other segments.

What else? The weight is just fine, and those two batteries account for lots of it!

Unresponsive Windows 10 laptop: automatic disk optimization

I have a 3-years old laptop at home (bought February 2015). HP, quad-core A8 CPU, 4 GB RAM, 5400 rpm HDD (slowish, I know). It came with Windows 8, then upgraded to 8.1 and finally to 10. It is used for browsing the Net and simple text and image processing. PCB design and Arduino coding, when needed. It has become less and less responsive without us adding new software. The symptom is the hard-disk being used 100% of the time without any apparent reason.

The problem lies in the user, or better, in the way we use it. This laptop gets used in the dark silent sleepy quarter-of-hours that lay between housekeeping chores (doing the dishes, getting ready for the next workday, ...) and hitting the pillow. That's between 23 and past midnight.

As far as I could understand, on Windows 10 I can define a "let me full control of the computer" time range, up to 12 hours long, but it doesn't span across midnight. Instead we need to be able to use it quickly, for 15-20 minutes in late evening, while Windows apparently does all its chores rendering the machine unresponsive.

So, before throwing it away in despair I wanted to try the move to Linux (Ubuntu): I've been using it as my main operating system at work in a 99% Microsoft-based environment, besides I know what I'm doing. Once I confirmed that everything apparently works booting off a USB drive, I needed some HDD space to install Ubuntu.

The correct procedure consists in shrinking the HDD though a standard Windows tool. That's when I saw the red alert light: Windows (10) is configured to do an automatic disk optimization once a week. On a 500 GB HDD full at 75% it could take half an hour. But what if in the meantime the antivirus updates, you use the computer and you shut it down because you're fed-updone? The whole process takes longer and I think it goes into an endless disk optimization, because it almost never completes.

Regardless this discovery I will install Ubuntu (16.04 LTS) on this HP 15-g005nl because there's no software keeping us tied to Microsoft Windows operating system.

Meanwhile, if you solved Win10 unresponsiveness due to 100% disk usage, let me know in the comments (and stay in-topic, I won't allow Win vs Linux flame or bad words about any of the products: tested solutions only).

Modding a DIY USB Power Bank case with HT4936S

Since I had some spare 18650 Li-Ion cells laying around, I decided to fit them in a couple of empty power bank shells sold online for few €/$. Today I will report on what looks like to be a copy of Xiaomi 10400 mAh power bank.

Knowing my pre-charged cells were relatively good, I suspected something was wrong when right after the first few seconds using the DIY USB power bank to charge my smartphone the status LEDs reported 75% or even 50% of charge remaining.

A quick inspection with the DVM after recharging the power bank revealed that my four cells were all measuring 4.0V, while I would and could charge them up to their nominal 4.2V voltage. Time to troubleshoot the electronics.

With the help of a magnifier I could identify the all-in-one controller chip as HOTCHIP HT4936S (Company website). Datasheet can be found online (7 pages in Chinese) and translated into English with an online service. The datasheet is far from being comprehensive, but the device isn't too complex either. The HT4936S chip takes care of charge and discharge control, voltage booster to 5V/1A, 4 status LEDs, a control button and, as every Respected Chinese product, a flashlight (press twice the button to activate - if the flashlight LED is present - twice again to turn off).

A note about the flashlight. This power bank does not mount the flashlight LED, but I tested the double click on another "shell" with unmarked controller chip and it works.

At first sight the circuit on the PCB is very similar to the datasheet sample application. Good. Some components are missing from their pads. Not so good.

Problem #1: cells are charged up to 4.0V only.
According to the datasheet, the HT4936S can be configured to stop charging either at 4.20V or 4.35V: while pin 1 is left open it should charge cells to 4.20V, if grounded it goes up to 4.35V. Why does it stop at 4.0V then? Shorting together R1 pads will ground pin 1 and set the target voltage to 4.35V. It works, and it keeps charging. I stopped it at 4.25V/cell so they don't get damaged.

Solution #1.
While this mod forces the IC to charge cells up to 4.35V, it moves "upwards" the discharge detection, therefore the chip considers them depleted at 3.6V thereabout. There is still the doubt why it stops at 4.00V instead of 4.20V in the default configuration. So this solution is not the final answer.

Problem #2: ground connection intrinsic resistance.
The ground connection is made of a single long wire that is wind-up to make the four contact springs: unless you insert a pre-made 1S4P pack, negative terminals of your 4x 18650 pieces won't be at the same potential (under load). I measured 100 mV drop between GND pad on PCB and the farthest cell during full current recharge (about 800 mA input). A solution could be to add a thick wire/strip that connects together all negative terminals.
In case you wonder, this extra resistance does not seem to be the culprit of Problem #1 since it keeps charging up to 4.35V.

Solution #2.
There is no much room in the power bank case to shorten the ground path. So ... live with it.


I wonder if off-the-shelf power banks suffer from the same design errors. Probably not. I will consider these as an energy source when no other solution is available. On the other hand I will not let "waste" 4 cells in there, but rather leave just a couple.

Count-up timer with TIL305 - optimizations

Looking at the 2x TIL305 timer I tried to imagine an even simpler version with lower parts count: could a single ATtiny2313 drive both displays? This AVR comes in a 20 pin package: take out two for power and one for reset (needed for low voltage programming and anyway it has limited current drive capability). That leaves 17 I/O pins.

My current implementation uses 5 pins for columns, 7 for rows and 1 for the decimal point: 13 I/O in total. Another display needs 5 control lines, we're 1 short (17 - 13 = 4).

If the decimal point is left out math tells us that 17 (total) - 13 (one full TIL305) + 1 (decimal point) = 5. That's enough to drive two TIL305 with one ATtiny2313.

But there is no room for the microcontroller to interact with the outside world, especially to receive an input signal like the 1 Hz clock. There is a way out!

When looking at the TIL305 running in landscape mode, you realize that your 5x3 symbols take up a total display area of 5x(3+3) pixels: the middle column (row #4) is never used because it acts as symbol spacer!

So, as long as you want to use 2x TIL305 in landscape mode displaying 4 symbols drawn on a 5x3 dot matrix, an ATtiny2313 will suffice as driver. You can now drop the code that sets row #4, thus running the scan 1/7th faster!

Want to get rid of the external 1 Hz timebase? Upgrade to an ATmega48 (o 88, 168, 328) and generate the clock with a XTAL so that the timing obtained in software will be more accurate than the internal 4/8 MHz RC.

Count-up timer with TIL305 - built

Thanks to a pre-wired breadboard the circuit went together in an hour:

From left to right: most significant digits and ATtiny4313, least significant digits and ATtiny4313, CD4060 oscillator and divider. Decoupling capacitors on the Vcc supply rail (that must be used in digital logic circuits!) prevent detection of false 1 Hz input pulses.


The circuit in the picture is powered with a 18650 LiPo cell at about 3.8V and it has been running all night to test the firmware past 10 hours of operation.

At powerup is starts counting immediately. First only seconds are shown. Then minutes and seconds until 59' 59". Then it switches to what is shown in the picture. After 9h 59' you get HH MM.

The font was chosen as a trade off of readability vs cozyness vs current consumption. There is some MSD blanking too, but it doesn't last long.

Further improvements before looking for a suitable box: add the battery, a 1S LiPo (dis)charge controller, a power switch. Maybe some firmware improvement once the timer is shown to the public.

Besides as kitchen timer, it can be used as elapsed work(-day) timer ... that's where I will hopefully gather some extra feedback.

The firmware is written in BASCOM. I want to do some further optimisation before uploading the code to my github account. Binary is available on request as well as TIL305, if need too.

Count-up timer with TIL305 - design

After building few clocks and thinking of many more, I wanted to do something slightly different. The "new" idea came one evening while cooking, when I realised I never remember at what time the pasta (food) will be ready. "Was it 20:14 or 20:20?"

I am too lazybusy to set a countdown timer every time and since my short-term memory is not willing to cooperate I chose to try a count-up timer: a simple device that shows elapsed time.

For this task I wanted to use TIL305 5x7 dot-matrix LED displays made in 1970's. Two challenges:

1. they need an appropriate driver
2. they are very current hungry

I had already solved the driver problem by reproducing someone else's idea to adapt an ATtiny2313 microcontroller. You need one ATtiny every TIL305. Cost-wise in 2018 this solution is probably cheaper than using obsolete unobtanium driver chips.

Texas TIL305, Monsanto MAN2A, ...

Since I want my timer to be battery powered, at about 38 mA per 2313&305 pair, I cannot use too many displays. Four digits is the minimum useful IMO ("mm:ss" and "hh:mm" for longer runs), but another idea comes to the rescue: use smaller font, 5x3 dots vs 5x7 and show two digits on each TIL305. This approach has another advantage: limit the required wiring (I won't be doing a PCB!).

How about the timebase? ATtiny's will not synchronise, they will receive both the same external timebase signal and rely on their internal clock to run the code. I could have had one master driver generating the pulse for others, but I am not sure on the internal RC calibration and on longer runs the difference vs 1 pps could be unacceptable. So there will be a 32.768 kHz oscillator and divider with CD4060 chip (because it accepts a wide range of supply voltage).

Old vs modern high voltage capacitors

Side by side comparison of the size of 500+ V capacitors, same values. 1955 vs 2018!

Well, the dielectric is different!

First MW reception on the RD-160

So, with the lamp limiter just built I could dare to power the Radiomarelli RD-160 medium wave receiver.

I scrounged home for filament light bulbs and I located 40W, 60W and 100W.

I stared with 40W, that lit up and nothing came out of the receiver. Maybe I wasn't patient enough, or its rating is just too low.

One step further to 60W and ... after the first bright flash, the lamp dim dim dim .... until some noise comes out of the receiver! The residual light is very warm and pleasant too! It works!! It receives RAI1 broadcast on 999 kHz and nothing else.

Problems noted:

1. When the tuning capacitor blades go towards full mesh the local oscillator stops: there's a short.
2. The volume control crackles
3. Knobs don't hold tight on the shaft because their plastic is cracked near the screw

Not bad at all! It is electronically fine! Now I will replace capacitors.

My poor man's AC limiter

Reading around the Net, everyone mentions that powering old valve circuits under restoration should be done gradually by applying a lower voltage and observing what happens. That's until all capacitors are confirmed to be working (or replaced) and transformers are not shorted. This is accomplished with a tapped or variable autotransformer, that I don't have.

One solution would be to wire back-to-back two step-down transformers, like 220-12 and 220-24 that should provide 110V if connected 220-12-24-220. Of course, I don't have such a selection of old-school transformers.

The "poor man's AC voltage transformer" trick comes to the rescue: connect a filament light bulb in series with the D.U.T. and observe what goes on. It will not allow to control the actual voltage, nor to change it on the fly, but it will limit the current through D.U.T. thus reduce the risk of smoke/fire/damage. Lower wattage bulbs provide lower voltage to the D.U.T. and let less current pass through, therefore less damage is done in case of troubles.

As I understand it, the principle is that a fault in the D.U.T causing excess current draw will be tamed by the light bulb which in turn will light up and provide visual feedback of a problem. Begin with lower wattage and replace with higher, step by step, as the home stock of old filament light bulb allows.

In order to build a "safe" transformer device I decided to modify a power strip: it has a power cord, outlets and an embedded switch. The mod is to wire in series two outlets, the first being used to plug the light bulb (installed in an existing lamp, for example), the second for the D.U.T. A third socket, left in parallel with the second, can be used to measure AC voltage across D.U.T.

The light bulb limiter, finished.

Things to do and remember:

• AC mains is lethal;
• add visual signs to the power strip that it is NOT a power strip anymore;
• mark sockets with their intended use;
• filament light bulbs get hot, so do not hurry to change it for the next!

If all goes well, the first troubleshooting session will not trip my home's AC protection.

Time travel to 1956: Radiomarelli RD-160 Amico

I finally did it: I bought a valve AM radio.

Radiomarelli RD-160 "Amico".

It is a Radiomarelli RD-160 "Amico" built on or after 1956. It was sold as non-working for "just" 5 € after nobody bought it for months. No wonder... In the worst case, individual parts are worth more (if bought individually, for other projects, not on the current market).

The first, off-grid, check reveals that the case is clean whereas cracked (it's plastic, can be glued together). The circuit is relatively clean and not rusty, all valves are present. The audio transformer is a bit loose and has one suspicious floating wire. The volume knob turns indefinitely and doesn't click on/off. The tuning knob runs the indicator bar just fine. Looks like a good troubleshooting playground and, at worst, I will retrofit it with some sort of Internet-radio receiver (not my original idea).

Can't get along with AMS1117 regulator!

There is something wrong between me and AMS1117 linear voltage regulators: I keep frying them! It first happened on two adapters for breadboard, but in that case I suspected I initiated some self-oscillation by using input leads of different length (comments on the same blog post explain other causes, worth reading!).

Then I blew it on a Arduino Nano clone, those that also mount the CH340 USB chip. Maybe I did something wrong, I don't remember. That circuit is running with a 7805 now, and gets pretty warm.

Last event was last night, when testing a PWM voltage booster with the same Arduino Nano clone board ("same batch"). I fed 12V the board and it all became quite warm. Since I wasn't getting the PWM output I resorted to powering it via USB, but this time I used an inline USB tester to see the current consumption: 150 mA. That sounded wrong for a board that was doing nothing. Once I sorted out the firmware to produce PWM, I reconnected everything on 12V and I didn't get any output, nor I could talk to the board over USB. Basically dead. So I phisically removed the AMS1117-5.0 with pliers (I began unsoldering, but it was taking too long). Result? The board still drains 110 mA and it is not recognised on USB. To the recycle bin.

NEXT!

Well, this time I will build my own 5V regulator with the trusty 7805.